The 1783–1785 A.D. Laki-Grímsvötn eruptions II: Appraisal based on contemporary accounts

Abstract

The 1783–1784 Laki eruption along with intermittent explosive eruptions at the subglacial Gríms-vötn central volcano, represents a two-year-long volcano-tectonic episode within the Grímsvötn volcanic system , from June 1783 to May 1785. The Laki eruption is the second largest basaltic flood lava eruption in historical times (after the 934–940 A.D. Eldgjá event) and its consequences were disastrous for Iceland. Detailed eyewitness accounts of the Laki-Grímsvötn eruptions exist and are here analysed in order to reconstruct the sequence of events as accurately as possible. This information is catalogued and critical eruption phenomena , such as the timing and nature of seismicity, explosive activity and mechanics of lava flow emplacement is evaluated. The results show that the 1783–1785 activity on the Grímsvötn volcanic system included at least 14 eruption episodes. Ten of these episodes were associated with the eight-month-long Laki eruption and linked to rifting and gradual lengthening of the erupting fissure. Each eruption episode began with an earthquake swarm, leading into a vigorous explosive eruption on a new fissure segment that was followed by sudden increases in outflow of lava. Concurrent explosive eruptions were reported at the Grímsvötn volcano during four of these episodes. The remaining four eruption episodes were confined to the Grímsvötn volcano, which remained active until May 1785. Evaluation of descriptions of flow phenomena in conjunction with new field observations of flow structures indicates that endogenous growth (i.e., insulated transport and inflation) was the characteristic mode of flow emplacement.

Full text

Reviewed research article
The 1783–1785 A.D. Laki-Grímsvötn eruptions II:
Appraisal based on contemporary accounts
Thorvaldur Thordarson1,2, Guðrún Larsen2, Sigurður Steinþórsson2and Stephen Self3
1) Department of Geology and Geophysics, University of Hawaii at Manoa, Honolulu Hawaii
2) Science Institute, University of Iceland, Reykjavík, Iceland; moinui@soest.hawaii.edu, torvth@hi.is
3) Department of Earth Sciences, The Open University, Milton Keynes MK7 6AA, UK
Abstract –The 1783–1784 Laki eruption along with intermittent explosive eruptions at the subglacial Gríms-
vötn central volcano, represents a two-year-long volcano-tectonic episode within the Grímsvötn volcanic sys-
tem, from June 1783 to May 1785. The Laki eruption is the second largest basaltic flood lava eruption in
historical times (after the 934–940 A.D. Eldgjá event) and its consequences were disastrous for Iceland. De-
tailed eyewitness accounts of the Laki-Grímsvötn eruptions exist and are here analysedin order to reconstruct
the sequence of events as accurately as possible. This information is catalogued and critical eruption phenom-
ena, such as the timing and nature of seismicity, explosive activity and mechanicsof lava flow emplacement is
evaluated. The results show that the 1783–1785 activity on the Grímsvötn volcanic system included at least 14
eruption episodes. Ten of these episodes were associated with the eight-month-long Laki eruption and linked to
rifting and gradual lengthening of the erupting fissure. Each eruption episode began with an earthquake swarm,
leading into a vigorous explosive eruption on a new fissure segment that was followed by sudden increases in
outflow of lava. Concurrent explosive eruptions were reported at the Grímsvötn volcano during four of these
episodes. The remaining four eruption episodes were confined to the Grímsvötn volcano, which remained active
until May 1785. Evaluation of descriptions of flow phenomena in conjunction with new field observations of
flow structures indicates that endogenous growth (i.e., insulated transport and inflation) was the characteristic
mode of flow emplacement.
INTRODUCTION
The 1783–1784 A.D. Laki eruption, which took place
on a 27km long volcanic fissure in the Síða high-
lands of south Iceland (Figure 1), is the second largest
basaltic flood lava eruption (after the 934–940 Eld-
gjá event) in historic time (Thordarson and Self, 1993;
Thordarson et al., 2001). It represents the main vol-
canic event of a two-year-long (May 1783-May 1785)
volcano-tectonic episode within the Grímsvötn vol-
canic system, which featured intense earthquake ac-
tivity and eruptions at the Grímsvötn central volcano
in addition to those at the Laki fissures. The conse-
quences of this activity were disastrous for Iceland
and the atmospheric and environmental effects caused
by the emissions from Laki were felt over large por-
tions of the Northern Hemisphere (Traumüller, 1885;
Thoroddsen, 1914; Thorarinsson, 1979; Sigurdsson,
1982; Steinthorsson, 1992; Wood, 1992; Thordarson
et al., 1993; Fiacco et al., 1994; Thordarson, 1995;
Grattan and Brayshay, 1995; Grattan, 1998; Thordar-
son and Self, 2001; 2003).
The events of the Laki-Grímsvötn eruptions are
documented in many contemporary Icelandic chron-
icles (e.g. Einarsson et al., 1984; Thordarson, 1990,
1991). This information has been used by volcanolo-
gists to infer the course of eruptive events as well as
JÖKULL No. 53, 2003 11

Thordarson et al.
Skaftá
Núpsvötn
Kú afljótð
Mýrdalur
18°
19°
Mýrdalsjökull
Katla
64°
Grímsvötn
volcano
Vatnajökull
Skeiðarársandur
Öræfi
Grímsvötn caldera
Laki lava flow
Laki cone-row
20 km
Fljótshverfi
Hverfisfljót
Síða
highlands
Síða
Síðujökull
Skaft rjökullá
P
V
O
He
T
EVZ
NVZ
WVZ
Kj
Hu
H
RSISZ
S
Sk
Sf
Hr
Vík
K
L
N
Landbrot
Skaftártunga
Álftaver
Meðalland
ICELAND
Figure 1. Simplified map of the study area showingthe position of the Laki vent system within the Grímsvötn
volcanic system (thick broken line) and the location of the Fire districts, the areas most affected by the Laki
eruption. District boundaries are shown as broken-dotted lines. Also shown are other districts, the ice caps and
outlet glaciers mentioned in the text, along with position of the Grímsvötn andKatla volcanoes. The scarp re-
ferred to in the text is shown as hachured lines. The location the farms Kirkjubæjarklaustur (K) and Prestbakki
(P) is indicated by filled squares. L and N designate the position of Lómagnúpur Mountain and Núpsstaður
farm. Inset shows the location of rift zones in Iceland; box is area of figure and filled circles indicate places
elsewhere in Iceland that are mentioned in the text. Abbreviations are as follows: H, Hafnarfjörður;He, Hekla;
Hu, Húsafell; Hr, Hrútafjarðará River; Kj, Kjölur; O, Oddi in Rangárvellir; R, Reykjanes; S, Selvogsheiði; Sf,
Skjálfandaflót River; Sk, Skagafjörður; T, Tindfjöll and Þórsmörk; V, Vestmannaeyjar; WVZ, Western Volcanic
Zone; EVZ, Eastern Volcanic Zone; NVZ, Northern Volcanic Zone; SISZ, South Iceland Seismic Zone. – Ein-
faldað yfirlitskort af Vestur Skaftafellssýslu og nágrenni ásamt helstu örnefnum. Innsett kort sýnir staðsetningu
rannsóknarsvæðisins.
12 JÖKULL No. 53, 2003

The 1783–1785 Laki-Grímsvötn eruptions
to evaluate the mechanics of the Laki eruption (e.g.
Thoroddsen, 1879, 1925; Helland, 1886; Thorarins-
son, 1967, 1969, 1984; Thordarson and Self, 1993).
The accuracy and reliability of these accounts vary
greatly, but when evaluated critically and corrected
for inconsistencies (see Thordarson, this issue), they
are a rich source of information about the course of
events. They also provide information about eruption
processes that otherwise would be unobtainable.
Here we present a comprehensive compilation and
a critical analysis of the information on the course of
eruptive events in 1783 to 1785 as revealed in the
contemporary chronicles, which are published here
for the first time in English. We also present new
evaluations of key eruption parameters derived from
this analysis. This includes a re-examination of the
source, timing and nature of seismic swarms, ex-
plosive activity, and lava surges and their inference
for number of eruption episodes during this volcano-
tectonic episode on the Grímsvötn volcanic system.
We have also evaluated the characteristic modeof lava
flow emplacement using information from the eyewit-
ness accounts in conjunction with new field observa-
tions. An extended list of references is included and
English translations of the original text used in the re-
construction presented here are given as quotations in
the appendix.
GEOGRAPHICAL AND GEOLOGICAL
SETTING
GEOGRAPHY
Geographic features and place names are commonly
used in the contemporary accounts as reference points
for positioning the phenomenon that the authors are
describing. Thus to fully comprehend the importance
of the descriptions it is useful to havesomeknowledge
of the local geography.
The Laki fissures and lava flow are located
in South Central-Iceland in theshire of Western-
Skaftafell (Figure 1). A distinct topographical feature
in this area is a scarp, an old sea cliff, which trends
across the central part of the area from WSW to ENE.
Cultivated areas are all located along this scarp or on
the Holocene outwash plain to the south and consist of
six districts: Álftaver, Meðalland, Skaftártunga, Síða,
Landbrot and Fljótshverfi (Figure 1). These districts
were most severely affectedby the Laki eruption and
thus often referred to as the Fire districts. North of the
scarp is the Síða highlands, bounded on three sides by
the Skaftá and Hverfisfljót rivers. The highlands fea-
tureSW-NEtrendinghyaloclastiteridgesseparatedby
low-lying wetlands. A few gorges and smaller val-
leys trend northward from the main scarp into the Síða
highlands, including those of the Skaftá and Hverfis-
fljót rivers which now are largely filled by the Laki
lava (Figure 2). The area around the Laki fissures was
boggy before the eruption and the remains of these
bogs are preserved along the edges of the lava-filled
Úlfarsdalur and Varmárdalur valleys. Northeast of the
Síða highlands is the Vatnajökull ice cap where the
subglacial Grímsvötn central volcano is located (Fig-
ure 1).
GEOLOGY
The Laki cone-row is located within the Grímsvötn
volcanic system which consists of the Grímsvötn cen-
tral volcano and an ∼100 km long and ∼15 km-wide
volcanic fissure swarm (Figure 1). The Grímsvötn
volcanic system is tholeiitic in character and is esti-
mated to have erupted between 50–55km3of magma
during the Holocene (Jakobsson, 1979). The volume
of magma erupted in the ice-free region of the sys-
tem is 21.5km3, of which 15.1 km3(or 65%) were
extruded by the Laki fissures in 1783–1784. The
Grímsvötn volcano has erupted at least 40 times since
1598AD with an average repose period of 10–15
years. The most recent eruptions occurred in 1983,
1996 and 1998 (Thorarinsson, 1974; Jóhannesson,
1984; Guðmundsson et al., 1997; Larsen et al., 1998;
Sigmarssonet al., 2000). A noticeable pause in the ac-
tivity at Grímsvötn volcanooccurs from 1785 to 1823,
the 38 years following the Laki eruption (Thordarson
and Self, 1993).
LAKI FISSURES AND LAVA FLOW
The Laki vent system is 27km long, extending from
Úlfarsdalur Valley in the west towards the tip of Síðu-
jökull glacier in the east (Figure 2). It consists of 10
northeasterly trending en echelon volcanic fissures,
which together host more than 140 vents (Figure 3).
JÖKULL No. 53, 2003 13

Thordarson et al.
3 km
contour interval 100m
N
672
833
818
614
S uj kullíð ö
Bl nguræ
Innri-
Eyrar
Ytri-
Eyrar
Galti
Hn taú
Laki
Varm rfellá
Kaldbakur
Thver rfjallá
Geirlandshraun/hei ið
Skaftá River gorge
Hverfisflj t River gorgeó
Hverfisflj tó
Lei lfsfellðó
Úlfarsdalur
Úlfarsdalssker
Lauffell
Bl gilá
Ámundabotnar
St rikr kuró ó
Hrossatungur
Hraunsendi
Lambatungur
Kanafj llö
Skaft rdalurá
Sk lá
Hei ið
Hunkubakki
Holt
Ytri-Dalb ræ
Kirkjub jarklausturæ
Prestbakki
M rtungaö
Sy ri-Steinsm rið ý
Efri-Steinsm riý
Lj tarsta iró ð
Orustuhóll
B landú
Hvammur
Botnar? H lmar?ó
Flj tar?ó
Hnausar
H lmasel?ó
Sk larstapiá
Stapfoss
Nes
Brandaland
Á
Árfjall
Ásar
Skaftá
Geirlandsá
Skaftá
K aflj túð ó
100
100
100
200
200
300
400
648
648
506
535
732
688
517
373
200
300
300
300
400
500
600
600
500
500
400
400
K lfafellá
Thverá
N parúHvoll
Ytridalur
Eystridalur
K lfafellshei iá ð
Sk larfjallá
Kringlur
Kambar
Uxatindar
Eldgjá
Hellisá
Dj páú
Kambavatn
Lambavatn
Miklafell
Varmardalur
Langisj ró
Rootless cone groups
Laki lava flow
Laki cone-row
7b.
7d.
7a.
7e.
7f.
7c.
lakes
glacier
63 40’
o
18 0’
o
64 0’
o
18 30’
o
Figure 2. Map of the Fire districts and Síða highlands showingthe regional topography,relevant landmarks and
place names, along with distribution of vents, lava flows, and rootless cone (pseudocrater) groups. Note that
location of farms destroyed by the lava is shown as it was prior to 1783. The pen circles and labels 7a–7f show
the locations of the photographs in Figure 7. Modified from Thordarson and Self (1993). – Örnefnakort af
Eldsveitunum og Síðumannaafrétti. Hringir eru dregnir um staði sem sýndir eru á 7.mynd.
14 JÖKULL No. 53, 2003

The 1783–1785 Laki-Grímsvötn eruptions
Skaftá
kilometers
01.0
Laki
10
10
a
Ulfarsdalssker
Ulfarsdalur
Skafta
Hnuta
Blaengur Innri-Eyrar
Varmardalur
TC-1
TC-2
Lambavatn
Laki
Varmarfell
S
i
d
u
j
o
k
u
l
l
N
Hverfisfljot
1098
7
65
4
3
21
0
8 June 11 June 14 June
27 June 9 July 29 July 28 August? 7 September? 26 Sept?
24 October?
510 15 20 25 27km
Tephra
Lava
Other features
Cone and crater outlines
Lambagigar cone row
Individual fissures (1-10)
Scoria and spatter cones
Tuff cones
Proximal tephra sectors
Lava channel
Lava flow
Rootless cone groups
Laki graben
Older faults
Bedrock
b
..
Figure 3. Map of the Laki fissures: demonstrating the main features of the cone-row and surrounding lavas.
TC-1 and TC-2 indicate the tuff cones SW and NE of Laki, respectively. Extent of individual fissures, labelled
1 through 10, shown schematically. Cones on fissure 3 (arrow) where first visited and described by Sveinn
Pálsson in 1794. – Lakagígar, gjóskukeilur, gígaraðir og sprungur. Örin bendir á gígana sem Sveinn Pálsson
kannaði og lýsti 1794.
Typically, each fissure is delineated by a row of scoria
and spatter cones, although two tuff cones on fissures
4 and 6 interrupt this pattern. Mt. Laki is an older
subglacial volcanic edifice, which coincidently sub-
divides the vent system into two almost equally long
segments (Figures 2 and 3). These segments, how-
ever, should not be equated with the terms Útnorðurs-
gjá and Landnorðursgjá used by Steingrímsson in his
accounts because he did not know this natural sub-
division of the vent system at the time of the erup-
tion. These terms simply refer to on which side of Mt.
Kaldbakur the vent activity was visible from the Síða
district (Figure 4).
Lava produced on the vents to the southwest of
Mt. Laki flowed south over the western part of the
Síða highlands in two main lobes; one confined to
the Skaftá River gorge and the other to the Varm-
árdalur and Hellisá river valleys (Figure 2). As the
lava emerged from the Skaftá River gorge, it spread
out onto the cultivated lowlands of Síða, Landbrot
and Meðalland. The vents to the northeast of Mt.
Laki issued lava to the south and north away from the
fissures. Lava flowing to the south advanced down
the Hverfisfljót River gorge, a distance of 25km, be-
fore reaching the lowlands of Síða and Fljótshverfi.
Lava flowing to the north spread over the outwash
JÖKULL No. 53, 2003 15

Thordarson et al.
Laki cone-row
Grímsvötn
Laki lava flow
20 km
Skaftá
18°
64°
Grímsvötn
Vatnaj kullö
15-25o
P
Ú
L
NÖ
K
Laki fissures
Figure 4. Map showing the relative position of the three eldgjás mentioned by Steingrímsson accounts as it
would be when viewed from Prestbakki (P) or Kirkjubæjarklaustur (K). The double arrows point towards Út-
norðursgjá (Ú) and Landnorðursgjá(L), respectively and the dot represents the peak of Mt. Kaldbakur. Broken
lines show the sightings of the Austur-Landnorðursgjá(Grímsvötn) as described in the contemporary accounts:
N, Núpstaður; Ö, Öræfi. The vectors (inset) show the angular offset in sense of directionsin the Síða district at
the time of the Laki eruption. – Innbyrðis afstaða eldgjánna þriggja, Útnorðursgjá (Ú), Landnorðursgjá (L) og
Austur-Landnorðursgjá (Grímsvötn), sem Jón Steingrímsson vitnar oft til í lýsingum sínum af Skaftáreldum.
16 JÖKULL No. 53, 2003

The 1783–1785 Laki-Grímsvötn eruptions
plain of the Skaftá River in front of Skaftárjökull be-
fore entering the upper reaches of the Skaftá River
gorge. The total magma volume produced by the
Laki eruption is 15.1 km3±1km3, thereof 14.7 km3
was erupted as lava and 0.4km3as tephra (Thordar-
son and Self, 1993).
CONTEMPORARY CHRONICLES
The contemporary chronicles discussed by Thordar-
son (this issue) provide valuable information about
the Laki eruption and are used here to reconstruct the
course of events during the 1783–1785 Laki-Gríms-
vötn eruptions. This information has been catalogued
according to the processes it describes and the cat-
egories are seismicity, explosive activity, lava flows,
and sense of directions in the Fire districts (Thord-
arson, 1990, 1991). Tables A to D in the appendix
containquotationscontaining informationrelevant for
each of the four categories. Each table is organised
such that it relates to the course of events in chrono-
logical order.
For the sake of simplicity, the references in the
following sections are not to the original accounts, but
to the translated quotations tabulated in theappendix.
For example, reference (A1) refers to the quotation
with index number A1 in Table A. Also note that text
enclosed by [ ] are inserted as explanatory notes by
the present authors.
GEOGRAPHIC DIRECTIONS AND
LOCATION OF ACTIVE VENTS
The geographic features and place names are com-
monly used in the contemporary accounts as reference
pointsforpositioningthe phenomena being described.
The orientation of certain topographic features af-
fected the inhabitants’ perception of geographicdirec-
tions, and resulting in a 15◦–25◦counter-clockwise
offset of directions in the Síða and Landbrot districts
(Figure 4). It is important to consider these offsets to
fully comprehend the importance of the descriptions
in reconstructing of the course of events (Thordarson
and Self, 1988; 1993). Therefore, location of all rel-
evant landmarks referred to in the texts are given on
the maps in Figures 1 and 2.
Steingrímsson’s references to particular localities
give valuable information about sense of direction
in the Síða, Landbrot and Fljótshverfi districts (D1–
D16) at the time of the Laki eruption. The north-
ernmost boundary of these districts is a 200 m high
scarp, trending NE (065–075◦) in the Síða and Land-
brot districts and E (090◦) in the Fljótshverfi district
(Figures 1 and 2). The trend of this scarp was the con-
trolling factor on perceived directions and at the time
of the eruption, as the scarp was believed to run due
east-west along its entire length. This is confirmed
by contemporary maps made by Hólm (1771), Step-
hensen (1785) and Pálsson (1794), where the scarp
in the Síða district is shown trending east-west (see
Figure 1 in Thordarson, this issue). This results in
a counter-clockwise offset of directions given in the
Síða and Landbrot districts of the order of 15◦–25◦
(Table 1 and Figure 4). In the Fljótshverfi district, the
offset is minimal if any. Steingrímsson lived in the
Síða district and most of his directional information is
based on local conditions and landmarks in that dis-
trict. Therefore, the counter-clockwise offset in his
descriptions must be accounted for when the direc-
tions of reported features are evaluated. This offset in
the perception of geographic directions has prompted
re-evaluation of some major conclusions reached by
earlier studies (e.g. Thorarinsson, 1974, 1984; see
also next section).
The offset of geographic directions is consistent in
southerly and easterly directions, but is more variable
in northerly and westerly directions, with the great-
est degree of variation in westerly directions (Table
1). Many of Steingrímsson’s observations were made
from his home, the farm Prestbakki, but some were
also made from the farm Kirkjubæjarklaustur, where
the church was located (Figure 2). The relative po-
sition of these farms as perceived by the inhabitants
at the time differs considerably from their actual geo-
graphic relationship. A likely explanation for this dis-
crepancy is the change in the landscape that occurs
between these two farms. A large valley, trending
NW (320◦) cuts into the main scarp between Prest-
bakki and Kirkjubæjarklaustur. On the old maps this
valley is shown trending due north and Prestbakki
is shown NE to ENE (055–080◦) from Kirkjubæjar-
JÖKULL No. 53, 2003 17

Thordarson et al.
klaustur, where the true direction is N15◦E (Figure 2
and Figure in Thordarson this issue). Because of the
mistaken direction of the valley between the farms,
references to the geographic west in the vicinity of
Prestbakki are slightly north of west (or 280–300◦).
This effect can be seen in some of Steingrímsson’s de-
scriptions of tephra fall that he reported to have been
dispersed by westerly wind (D10). On the other hand,
Kirkjubæjarklaustur is situated right under the scarp
and at that location geographic west was inferred to
be parallel to the scarp (245–255◦) and therefore re-
sulting in a southward displacement of the reported
westerly directions at this location.
Table 1. Direction (in degrees) of geographic direc-
tions as they were sensed in Síða district at the time of
the Laki eruption and their offset compared to true ge-
ographic directions. – Frávik átta á Síðunni á 18. öld.
Cardinal directions True position1Offset2
rétt stefna frávik
Northwest 320◦to 325◦+5◦to +10◦
North — —
Northeast 345◦to 360◦- 45◦to -60◦
East northeast 20◦to 30◦-37.5◦to -47.5◦
East 65◦to 70◦-20◦to -25◦
Southeast 120◦to 135◦0◦to -15◦
South — —
South-southwest 181◦-21.5◦
Southwest 215◦to 225◦0◦to -10◦
West-southwest 245◦to 255◦-2.5◦to +7.5◦
West 250◦to 300◦-20◦to +30◦
1) Geographic directions in degrees (east of north) as esti-
mated from position of landmarks used as a point of refer-
ence by Steingrímsson in his accounts.
2) Shows deviation of the values in column 2 from the true
geographic directions. Clockwise offset is indicated by plus
sign, counter-clockwise by minus sign.
A considerable offset is also in reported northwest
and northeast directions, which after correction be-
come; NW = 325◦, NE = 345–360◦and ENE = 20–
30◦(Table 1). It is vital to keep this discrepancy in
mind when locating the active vents during the Laki
eruption using the contemporary accounts. In the fol-
lowing compilation of the eruption, geographic direc-
tions are given as stated in the contemporaryaccounts
(i.e., not corrected), unless otherwise noted. Degrees,
when given, refer to the actual geographic directions.
LOCATION OF ERUPTION VENTS
Reverend Steingrímsson (1783; 1788) identifies three
distinct eldgjás1in his accounts of the Laki eruption.
These eldgjás indicate the sites where he saw fire or
ash columns rising from the active vents from his
home, the farm Prestbakki. Two of these gjá’s clearly
refer to segments of the Laki vent system, whereas the
third gjá indicates vents that were situated a good dis-
tance to the east of the Laki fissures (e.g. Thordarson
and Self, 1988; 1993).
During the first days of the Laki eruption, the
fires were seen rising above the moors to the west of
Mt. Kaldbakur, in the direction of the Mörtunga farm
(320–325◦) when viewedfrom Steingrímsson’shome,
Prestbakki (Figure 2). These sightings indicate that
the fire fountains at the very beginning of the erup-
tion emerged from the southwestern most vents on the
Laki fissures near the Hnúta Mountain. As demon-
strated by Thordarson (this issue), these vents are sit-
uated right in front of the now lava-filled Úlfarsdalur
valley, as is stated in Steingrímsson’s accounts (B26).
He refers to these vents as either Útnorðursgjá (the
northwest vents) or Vesturgjá (the west vents), terms
that he later applied to all vent activity visible to the
west of Mt. Kaldbakur. On 29 July, eruption columns
were seen rising for the first time above the eastern
shoulder of Mt. Kaldbakur, in the direction of Mt.
Blængur (Figure 2). The line of sight to these vents
(i.e., fissure 6 on Figure 3) is 10◦to the west of north
but not to the northeast as indicated in the original
accounts. In his descriptions Steingrímsson refers to
this and other vent activity that revealed itself to the
1Eldgjá is a term that is often used in old Icelandic literature to describe erupting vents or sites. Literal translation of the term is “the
fissure of fire” and is clearly derived from the most common experience Icelanders have had with volcanic activity, a fiery display of
fountains emerging from a linear vent system or a fissure. Here, Steingrímsson uses the term more loosely or simply to indicate erupting
vents.
18 JÖKULL No. 53, 2003

The 1783–1785 Laki-Grímsvötn eruptions
east of Mt. Kaldbakur as either Landnorðursgjá (the
northeast vents) or Austurgjá (the east vents).
The fires at the third eldgjá were seen at least
10 times from July 1783 to May 1785 (Steingríms-
son, 1788). These fires were situated east-northeast
of Kirkjubæjarklaustur according to observers and re-
ferred to as Austur-Landnorðursgjá (the east north-
east vents). The actual location of this site has been
debated in the past, because no historical eruption
sites are known due east-northeast of Prestbakki or
Kirkjubæjarklaustur (Thoroddsen, 1925; Thorarins-
son, 1974; Thordarson and Self, 1988). However, the
following quotations taken from Steingrímsson’s de-
scriptions give a clear indicationon the location of this
eldgjá:
26 September 1783: “Another additional fire and steam
column, which had been seen before that summer, far east-
northeast of the Síða district, due north of Lómagnúpur
mountain, but north-northwest of Öræfi district, as it was
oriented by me and 8 others.”
January 1784: “Fires were still seen at the East-
northeast fissure. From Klaustur [i.e., Kirkjubæjarklaust-
ur], these fires were seen behind the northern end of Þverár-
fjall mountain. The inhabitants of Öræfi district had a better
view of these fires due to their location.”
Firstly, these descriptions show that this eldgjá
(i.e., eruption site) was separated from the other two
by a considerable distance to the east. Secondly, de-
spite the fact that Steingrímsson says the fires were
east-northeast of Kirkjubæjarklaustur, his directional
information indicates otherwise. The fires were seen
behind the northern end of Þverárfjall, showing that
they were actually located to the north-northeast (20–
30◦) of Kirkjubæjarklaustur (Figure 4). The only
way these fires could be observed from the vicinity
of Lómagnúpur Mountain was through the valley of
the River Núpsvötn, which is located along the east-
ern side of the mountain. This valley trends 3◦east
of north. The third orientation given by Steingríms-
son is from the Öræfi district east of Skeiðarársan-
dur, where the fires were seen to the north-northwest.
When these three cardinal references are plotted, it
is clear that the Austur-Landnorðursgjá eruption site
was at the Grímsvötn volcano (Figure 4).
COURSE OF EVENTS AS REVEALED
BY THE CONTEMPORARY ACCOUNTS
THE PRECURSORS OF THE ERUPTION
A farmer, Jón Eiríksson at Ljótarstaðir in the Skaft-
ártunga district observed the first sign of activity in
the area when he felt weak tremors in mid-May 1783.
These tremors largely went unnoticed until 29 May
when they increased to such an extent that the inhab-
itants in Skaftártunga district evacuated their homes
and slept outside in tents for their own safety. On 1
June strong earthquakes were felt over an area extend-
ing from Mýrdalur inthe west to Öræfi in the east, to
a distance of about 75 km from the Laki fissures. The
intensity and duration of these earthquakes steadily
increased through the first week of June or until the
onset of the eruption on 8 June (A1–3).
Signs indicating volcanic activity were seen in the
mountains or in the glaciers above the Fire districts in
the spring 1783 before the onset of the Laki eruption.
The crew of a brig claimed that they saw fires some-
where above the Fire districts, as well as southwest
of the Reykjanes peninsula, as they were passing the
southern coast of Iceland on their way to Hafnarfjörð-
ur in SW-Iceland (Figure 1, inset). It is not known
whether this brig was the Boesand, which arrived at
the beginning of May or the Torsken, which arrived
around 20 May. It was also reported that the inhab-
itants of the Fire districts often saw a bluish smoke
above the ground, which they related to volcanic ac-
tivity (B1). The location of the above-mentioned
activity is unknown, but may have been within the
Vatnajökull ice cap because the inhabitants of the Fire
districts did not specify the location of the site.
In summary: The Laki eruption was preceded by 3–4
weeks of seismic activity. It began with weak earth-
quakes in mid-May 1783 that increased in intensity
and frequency until the beginning of the eruption on 8
June. If the observation by the brig’s crew is correct,
this precursory phase may have resulted in a short-
lived eruption sometime in May 1783.
FIRST 45 DAYS OF THE ERUPTION
In clear and calm weather on Whitsunday, 8 June
1783, at about 9 a.m., people in the Fire districts
JÖKULL No. 53, 2003 19

Thordarson et al.
saw a black volcanic cloud rise up north of the high-
est mountains bordering the Síða district. It spread
out rapidly towards the south with tephra fall in Síða
and the western part of Fljótshverfi (Figure 5a), such
that it became dark indoors at midday. Black ash-
like dust and “hairs” (i.e., Pele’s hair) covered the
ground. Rainfall later that day contained so much vol-
canic dust that it looked like black ink (B3–11). No
tephra fall occurred in the adjacent Skaftártunga dis-
trict (B7). Later that day columns of fire were seen
both from Síða and Skaftártunga. According to Rev-
erend Einar Björnsson during his visit at Prestbakki,
three fires were seen northwest of the farm Mörtunga
(B8). Steingrímsson describes the same fires as nu-
merous columns of fire (B3). Jón Eiríksson saw one
column of fire on 8 June (B7). Tephra fall eased in the
afternoon on 8 June, which coincided with the arrival
of southerly sea breeze (B4). The following day the
plume was clearly visible from the Síða district and
still rising. A rainfall that evening was reported to be
from the volcanic cloud (B12).
The weather was cloudy on 10 June and acrid ash
laden rain containing Pele’s hair poured down most
of the day. The rain was accompanied by a sulfurous
stench and was so polluted that it afflicted the peo-
ple with dizziness and smarting in their eyes and skin.
The raindrops burned holes through leaves and pro-
duced spots on the skin of newly shorn sheep (B13–
15). On 11 June there was a strong wind from the east
with snowfall and drift (B17). Tephra-fall was only
reported from the moors above Skaftártunga (B19,
Figure 5a). On this day Jón Eiríksson at Ljótarstaðir
in Skaftártunga noticed two new eruption sites to the
northeast of the first one (B18). Between 11 and 14
June, Árni Thorarinsson saw three separate eruption
columns from his home, Oddi in Rangárvellir, over
100km to the west of the eruption site (B20). Earth-
quakes recommenced in the early hours of 9 June and
increased continuously until 11 June, which was the
last day of the continuous earthquakes (A4–7).
On 9 June, the flow of water in the Skaftá River
began to dwindle considerably and its channel dried
up on 10 or 11 June, apart from for the run off from
local tributaries (C3–7). A friend, who lived at the
farm Skál in the Síða district, sent a letter to Sveinn
Pálsson. The letter describes a trip undertaken on 11
June by him and a few other men that went woodcut-
ting in the Síða highlands along the eastern side of the
Skaftá River gorge. The riverbed had dried up so they
let the logs fall into the gorge. At the same time, they
noticed a huge steam cloud rise up from the north-
ern part of the gorge. Later that day they observed
a glowing lava stream, filling the gorge and advanc-
ing rapidly (the position of this lava front is shown on
Figure 5a). While they collected their belongings the
lava flowed over and burned all of the logs they had
cut down (C7).
On 12 June the first of five lava surges emerged
from the gorge onto the farmlands located on the
glacio-fluvial sandur plain in front of it (C8–12). At
the end of the day, the flow front had reached the
fields of the farm Á, about 3 km south of the mouth of
the gorge (C9; Figure 5a). On 14 June the front was
reported to be 20 fathoms (about 35 m)away from the
graveyard at the farm Skál (C13), some 30km south
of the first fissure to open.
On 13 June, the wind was from the southwest and
the eruption column rose high enough (i.e., »2 km)
that it could be seen from Selvogsheiði in southwest-
ern Iceland, 300km away (B23–24). On this day an
exploratory party climbed Mt. Kaldbakur to the north
of the Síða district to get a good view over the eruption
site. Steingrímsson notes that these men saw 20 fire
columns (fountains) in Úlfarsdalur Valley and 7 oth-
ers in the hillock further to east (i.e., Úlfarsdalssker)
(B25–26). The only lava stream seen by these farm-
ers was the one in the Skaftá Rivergorge. Earthquakes
recommenced on 13 June (A8–11) and on 14 June, in
calm weather and a light southerly breeze, a substan-
tial tephra-fall was reported in the Síða district (Fig-
ure 5b). This tephra fall contained such an abundance
of Pele’s hair that it rolled up into elongated bun-
dles blown around by the wind. Steingrímsson clearly
states that this tephra-fall contained much more of the
hairy material than previous ones. Acid and ash-laden
heavy rainfall, as previously described on June 10th,
occurred that night despite the southerly wind (B27–
29). The night before the 15th a second lava surge
came out of the Skaftá River gorge resulting in further
advance of the lava front onto to the plains in frontof
20 JÖKULL No. 53, 2003

The 1783–1785 Laki-Grímsvötn eruptions
the Síða highlands (C13–14, Figure 5b). Activity at
the fissures continued with vigour (B30–32). On the
17th of June the wind was from the north and it was
on this day that Steingrímsson saw from his home a
reddish blaze and sparks of fire (i.e., glowing magma
blobs) rising above Geirlandsheiði (C15). This activ-
ity has been shown to have resulted from rootless vent
eruptions in Laki lava in the vicinity of Mt. Leiðólfs-
fell, demonstrating that by this time the lava in the
Varmárdalur valley had begun its advance down the
Hellisá river channel (Figure 5b, Thordarson et al.,
1998).
On 18 June, the third lava surge emerged from
the Skaftá River gorge (C16–17) and was accom-
panied by earthquakes (A12–13). This lava surge
was enormous and the lava advanced rapidly in two
branches. One branch progressed to the east follow-
ing the channel of the Skaftá River, while the other
advanced rapidly down the channel of the river Mela-
kvísl reaching the Steinsmýri farms on 22 June, trav-
elling 15-16km in about 5 days (C18–20, C22–23;
Figure 5b). On 20 June, the farmers took another trip
to the mountains north of the Síða district to observe
how the eruption was proceeding. Three lava streams
were seen emerging southwards from the vents into
the Varmárdalur Valley. As these lava streams merged
they advanced southwards over the boggy pasture and
the fluvial plain of the Hellisá River (C21). The flow
continued down the riverbed of Hellisá until it joined
the earlier lava flow in the Skaftá River gorge just
west of Leiðólfsfell (Figure 2). Sources from Skaft-
ártunga indicate that lava in the Hellisá River channel
was flowing into the gorge by 24 June (C25, Figure
5c).
On 21 June, the wind was blowing from the east
along with sleet and snow. A considerable fall of
sandy tephra was reported from Síða while greyish
coloured sand was observed on thin boards and white
paper in Skagafjörður, Northern Iceland (B34–35;
Figure 1, inset). Easterly winds continued through
22–25 June (B36–40). Earthquakes were felt again
on 23–26 June (A14) and at the same time the lava
advanced at a steady rate from the gorge (C24). Ex-
plosive activity increased between 23–25 June, but
the plume was directed westwards by the wind and
may have caused tephra to fall north of Mýrdalsjökull
(B39–40, Figure 5c).
On 27 and 28 June, with a strong westerly wind,
a considerable tephra-fall occurred in the eastern part
of the Fire districts (i.e., in Síða and Fljótshverfi dis-
tricts; B41–43, Figure 5c). Steingrímsson’s descrip-
tion of the tephra fall on the 27th is as follows: “We
here in the Síða district saw a huge black sandy cloud rise up
from Vesturgjá, which twisted and spun over the mountains
forming bolsters [i.e., cumulus-shape plume]. This cloud
dispersed sandy ash containing some pumice-like lapilli
scoria over Fljótshverfi and the eastern part of Síða and pro-
duced by far the most voluminous tephra-fall in these dis-
tricts.” (B41)
This tephra-fall continued through the 28th , forc-
ing farmers in Fljótshverfi away from their homes and
to seek refuge in the settlements west of the Fire dis-
tricts (B43). The wind was easterly on 29 June and the
fourth lava surgeemerged from the Skaftá River gorge
that day (C26). From that time until 12 July, lava
flowed continuously out from the gorge, although at
pulsating rates, further enlarging the flow field (C27,
Figure 5c).
A considerable amount of tephra fell in Síða on 9
and 10 July, but it did not reach the Meðalland district
(Figure 5, B48–50). Increased activity was noted on
13 July, with thunder, lightning and earthquakes for
the next four days. The flow rate of the lava increased
and the last big lava surgeemerged from the gorge on
13 or 14 July and the subsequent advance of the lava
fronts completed the constructionof the western flow
field in front of the Síða scarp (C28–29, Figure 5d).
However, lava was seen flowing in the gorge for the
rest of the summer (C32, C36). The intensity of the
eruption appears to have increased between 18 and 20
July, such that livestock gathered in groups because of
seismic activity and earthquakes were reported on the
20th (A18–19, B52–53, C30–31). On 18 July, ash fell
in the Síða and Fljótshverfi districts and Steingríms-
son describes it as follows: “When these above described
phenomena had been going on the whole night and a large
part of the day, a sandy and muddy ash was dispersed over
the ground here so it blackened. This tephra-fall came from
east-northeast and differed from previous ones in colour and
touch. It came from another eldgjá [i.e., Austur-Landnorð-
JÖKULL No. 53, 2003 21

Thordarson et al.
18°
64°
18°
64°
18°
64°
Skaftá
Skaftá
Skaftá
Hverfisfljót
Hverfisfljót
Hverfisfljót
18°
?
June 21?
July 18
July 19
8 June
Mt. Laki
Mt. Laki
14 June
19 June?
11 June
11 June
12 June
12 July
~21 July
12 July
14 June
18 June
22June
24 June
24 June
27-28 June
25-26 June?
9-10 July
11-13 July?
64°
Skaftá
Tungnaá
Tungnaá
Tungnaá
ICELAND
Grímsvötn Grímsvötn
Grímsvötn Grímsvötn
Vatnaj kullö
Vatnaj kullö
Vatnaj kullö
Vatnaj kullö
?
?
?
?
?
Laki cone-row
Laki lava flow
shades of gray show
location of flow front
at the indicated dates
20 km
a) b)
d)
c)
Mt. Laki
Mt. Laki
Tephra dispersal direction
and area of tephra fall at
given dates
rivers and lakes
dried up river channels
Figure 5. a-g) Origin, timing, and dispersal of reported tephra falls in summer and fall 1783. The position of
the lava flow front as indicated in the contemporary descriptions at the given dates is also shown, depicting
the approximate growth of the lava flow field with time. – Uppruni, tímasetning og útbreiðsla gjóskufalls og
hrauns frá Lakagígum og Grímsvötnum sumarið og haustið1783.
22 JÖKULL No. 53, 2003

The 1783–1785 Laki-Grímsvötn eruptions
e)
18°
18° 18°
18°
64°
64° 64°
64°
Skaftá
Skaftá
7 September
September 14
29 July
Tungnaá
Tungnaá
Grímsvötn
Grímsvötn
Vatnaj kullö
f)
g)
Eruption columns in fall 1783
September 26
October 25
November 24
Eruption columns in 1784-85
Late January 1784
8 April 1784
Early May 1785
May 26 1785
h)
9 August
7 September
29 October
Mt. Laki Mt. Laki
Mt. Laki
Hverfisfljót
Hverfisfljót
Hverfisfljót
Hverfisfljót
Skaftá
Grímsvötn
Vatnaj kullö
Vatnaj kullö
Skaftá
Tungnaá
Grímsvötn
Mt. Laki
Vatnaj kullö
Figure 5. cont. h) Dates when eruption columns or other activity was noticed at the Grímsvötn volcano in 1784
and 1785. – h) Tímasetning þegar gosmökkur sást rísa uppfrá Grímsvötnum.
JÖKULL No. 53, 2003 23

Thordarson et al.
ursgjá] as later became evident.” (B52) [This tephra fall
originated at Grímsvötn volcano, see also Figure 5d].
On 19 and 20 July, this ash cloud was directed to-
wards the west by easterly winds, but to the northwest
on the 21st (B54–56). From that day: “no loud rum-
bling sounds were heard from the Vesturgjá, despite the fact
that fires were seen and lava continued to flow from it into
September.” (B56)
Fires were seen at the fissure to the end of July
and an intense fire-coloured glow was seen in the
sky on 24 and 25 July (B58, B60). Fine ash, de-
scribed to look like finely ground tobacco, fell at
the Kjölur mountain-route some time before 24 June,
about 250km north-northwest of the Laki fissures
(B59; Figure 1, inset). On the 28th rain containing
ash fell in Síða district and was accompanied by a bad
stench (B61).
In summary: The chronicles report seven occurrences
of tephra fall in the Fire district during the first 45
days of the eruption (Figure 5a-d). Five of these
tephra falls can be traced back to the Laki fissures,
whereas two (i.e., 21 June and 18–21 July) clearly
originated from a more easterly source, namely at the
Austur-Landnorðursgjá (= Grímsvötn volcano). The
chronicles also show that five of the tephra falls from
the Laki fissures were preceded by an increase in
earthquake activity and that each was followed by
a major lava surge emerging from the Skaftá River
gorge three to five days later (Table 2). The end of
this period also associated with marked change in the
eruption, as the focus of activity was shifted to sites
the northeast of Mt. Laki (see below and Figure 3).
COURSE OF EVENTS IN LATE SUMMER AND
FALL 1783
On 29 July, rumbles and a boiling sound were first
heard north of Mt. Blængur, accompanied by tremors
of similar intensity to those generated by the activ-
ity on Vesturgjá. Shortly thereafter, a “huge and
dreadful” volcanic cloud emerged from this site, dis-
persing sandy ash and pumice-like scoria over Fljóts-
hverfi and the eastern part of Síða (B62, Figure 5e).
Earthquakes continued and were felt intermittently
throughout August (A20–21). The water in the Hverf-
isfljót River began to warm up on 3 August when a
steam cloud was seen advancing down its gorge. The
river dried up the following day. The first lava surge
emerged from the gorge on 7 August and two days
later the lava had advanced 4km beyond the hill Orus-
tuhóll (C37–40, Figure 5e).
Intense rumbling and cracking was heard from
the new fissure between 30 July and 7 August and a
continuous boiling sound was heard until 23 August
(B63, B66, B68–70). Southwesterly winds kept the
eruption column away from the Fire districts for the
first two weeks of August, except for 6 August when
fine-grained tephra was dispersed over Fljótshverfi
(B67). Showers containing ash fell in the Síða dis-
trict between 17 and 23 August (B72). Some time be-
tween June and August the Laki eruption column was
seen from the farm Húsafell in western Iceland rising
above the eastern part of the Kaldidalur Valley (B71),
indicating column heights in excess of 9 km (Thordar-
son and Self, 2003). The lava emerged at a steady rate
from the gorge until 14August, but then the outflowof
lava dwindled considerably and remained so through
August (C42).
Earthquakes started again on 23 August and con-
tinued to the end of the month (A21). On 1 Septem-
ber a second lava surge emerged from the Hverfis-
fljót River gorge and in the following days a branch
of the lava advanced rapidly to the east towards the
farm Núpar (Figure 5f). The flow of lava was so in-
tense on 7 September, when Steingrímsson crossed
the Hverfisfljót sandur plain on that day all the rivers
and streams were dry (C43–44). In the Fire dis-
tricts the first week of September was characterised
by sandy rain, mist, thunder and lightning (B76).
Ash-fall was reported in Vestmannaeyjar on the
7th, lasting for 2–3 hours (B77, Figures 1 and 5f). The
third lava surgeemerged on 10 September andthe lava
flowed in between earlier lava (C45). September 14
came with an easterly wind and heavy ash-fall which
only reached as far west as the Geirlandsá River (B78,
Figure 5f). From this day until the 26th an intermis-
sion occurred in the flow of lava from the Hverfis-
fljót River gorge and the flow of water resumed (C46).
According to Steingrímsson, the Skaftá and Hverfis-
fljót rivers reappeared in their gorges at this time. The
glow of fire was seen at nightabove the mountainsand
24 JÖKULL No. 53, 2003

The 1783–1785 Laki-Grímsvötn eruptions
Table 2. Timing of earthquakes, explosive activity and lava surges during the Laki-Grímsvötn eruptions in
1783–1785. – Atburðarás og einstakar goshrinur Skaftáreldaog Grímsvatna 1783–1785.
Earthquakes Tephra Fall Lava surges Eruption episodes
jarðskjálftar öskufall aukið hraunrennsli goshrinur
19 May–8 June 8–9 June 12 June I
9–11 June 10–11 June 14 June II
13–19 June 14 June 18–20 June III
21 June3?
23–26 June
30 June 25–28 June 29 June IV
14 July 9-10 July 14 July
17–20 July 18–19 July318 July?2V
29 July?
1–3 August 29–30 July 7 August VI
23–31 August 2 Sept. VII
7 Sept.
14 Sept. 310 Sept. VIII
26 Sept. 26 Sept. 4IX
24–? Oct. 25 Oct. 325–29 Oct. X
2 Nov. ?
24 Nov. 24 Nov. 3XI
Jan. 17843XII
Feb. 17841
8 April 17843XIII
14–25 Aug. 17841
4–26 May, 17853XIV
1) These earthquakes were not directly linked to the Laki eruption or the volcano-tectonic episode at the Grímsvötn volcanic
system. See text for further discussion.
2) Steingrímsson’s description imply vigorous activity at the fissures on this day (i.e., B30), but does not directly indicate
increase in lava production and no lava surge was observed from the Skaftá River gorge.
3) Fires seen at or tephra fall from Austur-Landnorðursgjá (i.e., Grímsvötn volcano).
4) This lava surge never reached the lowlands, but increased lava production is indicated by cessation of flow in both
Hverfisfljót and Skaftá rivers.
from the farm Prestbakki it was observed to be most
intense in the lows between Kaldbakur and Lamba-
tungur in the northwest and Kaldbakurand Vothamrar
in northeast (B79).
Strong earthquakes were felt again on 26 Septem-
ber (A22). The emission of lava from the fissures in-
creased to such a degree that most rivers ceased flow-
ing, including the Skaftá and Hverfisfljót rivers (C47).
This was accompanied by, “An additional column of fire
and steam, which had been seen few times before [i.e., 18
July] far east-northeast from Síða, due north of Lómagnúp-
ur, but northwest from Öræfi district.” (B80, Figure 5g).
This activity continued at a diminishing rate until
24 October when intense earthquakes were felt in the
Fire districts and an acrid stench was emitted from the
east-northeast (A23). On 25 October, a huge column
of fire rose from the Austur-Landnorðursgjá and rain
containing sandy ash fell in the Síða district (B81–82).
The same day an intensive lava surge emerged from
the Hverfisfljót River gorge and in the following five
days the lava filled in the gap in between two earlier
flows (C48, Figure 5g). This was the last lava surge
JÖKULL No. 53, 2003 25

Thordarson et al.
to come out of the gorge. Throughout November,lava
continued to issue from the fissures but the new flows
were confined to the highlands (C49). Rainfall, con-
taining ash and sand occurred frequently in late Oc-
tober to November (B83). On 24 November, a strong
earthquake was felt in the Fire districts (A25), which
was followed by increased activity at Austur-Land-
norðursgjá, such that the fire column was seen to rise
above Mt. Kaldbakur (Figure 5g). In the beginning of
December the fire glow abovethe Laki fissures, which
so far had been seen every night, began to diminish
(B83). However, lava was still emerging from the fis-
sures and the bluish haze continued to hover over the
Fire districts (C49).
In summary: Three tephra falls are described in the
chronicles for this period of the eruption (Figure 5e
and f). Two of these tephra falls (29 August and 7
September) are traced back to the Laki fissures and
one (i.e., 14 September) is inferred to have originated
at the Austur-Landnorðursgjá. Also, distinctive ex-
plosive eruptions occurred at the Austur- Landnorð-
ursgjá on 26 September, 25 October and 24 Novem-
ber (Figure 5g). Similar to the occurrences early on in
the eruption, the explosive activity is closely associ-
ated with occurrences of earthquake swarms and lava
surges. This time the lava surges emerged from Hverf-
isfljót River gorge because of the northeast shift in the
location of the most active vents (Figures 2 and 3).
THE FINAL SCENARIO
A sulfuric stench filled the air in the Fire districts fre-
quently throughout 1784, especially when the wind
was blowing off the vent areas. This was also com-
monly noted in calm weather and on rainy days (B86).
The residents of the farm Skaftárdalur in Skaftártunga
saw fires at the Laki fissures until 14 January 1784.
According to Steingrímsson, no activity was observed
on the fissures after 7 February 1784 (C50).
The Austur-Landnorðursgjá continued to erupt
through the winter of 1784 (B86–87; Figure 5h).
Flash flooding (jökulhlaup) occurred in the Núpsvötn
(Súla) River on 8 April and at the same time a huge
black volcanic cloud rose from Austur-Landnorðurs-
gjá. Steingrímsson reports that ash from this erup-
tion site fell twice in the Fire districts, but most of
the time the plume was directed away from the dis-
tricts by southerly wind (B87). Steingrímsson’s de-
scriptions indicate that this activity lasted fora consid-
erable time, possibly throughout the spring of 1784.
Activity at Austur-Landnorðursgjá was observed off
and on through the year 1784 and in the winter 1785,
especially from the Öræfi district. The last day when
fires were seen at Grímsvötn was 26 May 1785 (B89).
Strong earthquakes were felt in the Fire districts
sometime in February 1784 and were described by
Steingrímsson as follows:
“After a long period of freezing and heavy winds which
lasted through January, strong earthquakes were felt again,
though different than previous ones. Some moved the earth
slightly up and down, others travelled like oceanic waves
from northwest to southeast. The frozen ground cracked
with the sound of thumps and claps.” (A26)
The cracks were most numerous in the Fljóts-
hverfi district and Steingrímsson notes that the spring
water in the Fire districts, which had been strongly
contaminated up to this time, tasted much better after
these earthquakes.
DISCUSSION
NATURE OF ERUPTIVE ACTIVITY AND
RELATED PHENOMENA
The Laki eruption began on 8 June 1783 and 7 Febru-
ary 1784 was the last day when activity was noticed
at the fissures. This day is taken to mark the end of
the Laki eruption and accordingly the eruption lasted
for 8 months. Despite the termination of activity on
the Laki fissures, activity continued at the Grímsvötn
volcano until 26 May 1785, indicating an ∼2 year du-
ration for this eruption episode on the Grímsvötn vol-
canic system (Thordarson and Self, 1993).
For almost 200 years the accepted view on the
progress of the Laki eruption was based on the idea
put forth in 1794 by Sveinn Pálsson (see Thordarson,
this issue). The view was that the eruption featured
two main eruption episodes, the first commencing on
8 June on the fissures to the southwest of Mt. Laki
and a second one that commenced on 29 July on the
fissures to the northeast of mountain (Figure 3). Both
episodes where thought to have begun with explosive
26 JÖKULL No. 53, 2003

The 1783–1785 Laki-Grímsvötn eruptions
activity that was followed by a longer-lasting phase of
quiet lava effusion (i.e., Helland, 1886; Thoroddsen,
1879, 1894, 1925; and Thorarinsson, 1967, 1969).
However, studies of the proximal tephra strati-
graphy have shown this view to be oversimplistic
and that the Laki eruption featured at least 10 dis-
tinct eruption episodes (Thordarson and Self, 1993).
Furthermore, the Grímsvötn volcano eruptedintermit-
tently from June 1783 through May 1785, adding at
least four eruption episodes to the Laki-Grímsvötn
event (Table 2). These results are substantiated by the
contemporary descriptions (Thordarson, 1990, 1991),
which provide strong evidence for the episodic nature
of the Laki eruption as indicated by repeated occur-
rences of earthquake swarms, explosive activity and
lava surges. The nature and association of these phe-
nomena, as revealed in the contemporary accounts,
will now be considered in more detail.
SEISMIC ACTIVITY
Three facets of the seismic activity associated with the
Laki eruption have importantimplications. Firstly, the
contemporary accounts only document earthquakes
that were strong enough to be felt by the inhabi-
tants at a distance of 30–80 km from the Laki fissure.
Secondly, they indicate that earthquakes occurred in
swarms that often lasted for days. Thirdly, these earth-
quake swarms commonly preceded or coincided with
the onset of explosive activity and increased lava pro-
duction at the vents (Table 2).
The evidence indicates that the earthquakes asso-
ciated with the Laki eruption were readily detected at
distance of >30km from source vents, implying that
their magnitude was ≥4 on the Richter scale (Brands-
dóttir, 1992). Consequently, rifting of brittle crust
rather than volcanic tremor (sensu stricto) is a more
likely source for these earthquakes.
Also, despite a significant difference in peri-
odicity, the episodic occurrence of intense earth-
quake swarms in association with increased activity
at the Laki fissures shows a strong resemblance to
the pattern observed in recent volcano-tectonic rift-
ing episodes in Northern Iceland (Bjornsson et al.,
1977; Brandsdóttir and Einarsson, 1979, Brandsdóttir,
1992). Consequently, it is logical to conclude that the
3–4 weeks of gradually increasing earthquake activ-
ity that preceded the Laki eruption mark the onset of
rifting within the Grímsvötn volcanic system and the
initial emplacement of the Laki feeder dyke. By the
same token, the subsequent earthquake swarms rep-
resent ongoing rifting that resulted in gradual north-
east extension of the erupting fissures (Thordarson
and Self, 1993).
The strong earthquakes felt in the Fire districts
in February 1784 are worthy of further note. These
earthquakes occurred towards the end or after ces-
sation of activity at the Laki fissures. Furthermore,
Steingrímsson describes these earthquakes as being
different from those clearly associated with the erup-
tion and it is very likely that these earthquakes were
not directly related to the activity at the Laki fissures.
It is interesting that strong earthquakes were also felt
on the Reykjanes peninsula in Southwest Iceland on
24 and 25 February 1784(Lievog, 1784). It is possible
that these earthquakes resulted from larger-scale dis-
locations on the South Iceland Seismic Zone (SISZ)
and may have been the precursors to strong earth-
quakes that occurred onthe SISZ in August that year,
causing significant damage to structures and buildings
in the region. Furthermore, it is difficult to believe
that this sequence of events (i.e., the Laki eruption and
subsequent earthquakes on the SISZ) is a random oc-
currence because not only do they show close associ-
ation in terms of time, but occur on structures that are
closely linked geologically. The SISZ is a complex
zone of faulting linking the overlapping sectors of
the Western and the Eastern Volcanic Zones and thus
it is conceivable that the earthquake activity on the
SISZ was triggered by the Laki-Grímsvötn volcano-
tectonic episode (Stefánsson and Halldórsson, 1988;
Gudmundsson, 2000).
EXPLOSIVE ACTIVITY AND TEPHRA
DISPERSAL
Although the tephra fall deposits only amount to 2.7%
of the erupted magma volume (Thordarson and Self,
1993), the explosive activity at the Laki fissures was
significant and should not be overlooked. The mag-
nitude of the explosive activity at Laki gains a new
perspective when it is acknowledged that the 0.4 km3
JÖKULL No. 53, 2003 27

Thordarson et al.
of tephra produced by the eruption is greater than the
cumulative tephra volume from all four 20th Cen-
tury Hekla eruptions. The Laki tephra deposit cov-
ered 7200km2within the 0.5 cm isopach. For com-
parison the area within the 0.5 cm isopach for the
1947 and 1991 Hekla eruptions is ∼2000km2and
485km2, respectively (Thorarinsson, 1976, Larsen et
al., 1992). The Laki eruption plumes also dispersed
tephra as far as the Faeroe Islands, Scotlandand main-
land Europe, implying that an area of ∼750,000km2
may have been affected by tephra fall from Laki. In-
formation on the timing and dispersal of tephra falls
provides valuable information on the source and na-
ture of explosive activity during the Laki-Grímsvötn
eruptions. This facet of the eruption has been anal-
ysed in detail by Thordarson and Self (1993) and thus
only the key results are reiterated here.
In the first 3 months of the eruption, the contem-
porary accounts report fifteen occurrences of tephra
fall in association with the Laki-Grímsvötn eruptions.
Eight of the reported occurrences refer to tephra falls
that directly affected the Fire districts and five indi-
cate rain out of tephra in the highlands above the Fire
districts (Figure 5). The last two are reports of tephra
fall elsewhere in Iceland, one near Vestmannaeyjar off
the south coast and another in Skagafjörður, Northern
Iceland (Figure 1, inset).
Eight of the reported tephra falls originated at the
Laki fissures (Figure 5a–f). On five occasions the
tephra was spread over the Fire districts, whereas on
three occasions the tephra fall was confined to the
highlands to the west and north of the fissures. Each
of these tephra falls are derived from different fissure
segment on the vent system and the tephra stratigra-
phy indicates a systematic northeast propagation of
the explosive activity with time. Furthermore, textural
studies of the tephra fall units that make up the proxi-
mal deposit show that at least five of the fall units were
produced by explosive activity of subplinian character
whereas two were formed by phreatomagmatic activ-
ity (Thordarson and Self, 1993).
Sightings of fire fountains on 8 June indicate that
at the onset of the eruption the activity was confined
to the vents on fissure 1 (B8; Figure 3). The tephra
that fell in the Fire districts on the first day was pro-
duced by explosive activity on this fissure (Figure 5a).
The two additional fire fountains noticed on 11 June
indicate a northeast extension of the vent system and
were linked to formation of fissure 2 (B18; Figure
3). The following night explosive activity on these
fissures produced a tephra fall that was dispersed to
the west-southwest toward the Mýrdalsjökull glacier
(Figure 5a). The tephra that fell in the Fire districts on
14 June contained noticeably more of Pele’s hair than
previous falls (B27–B29) and the tephra stratigraphy
shows that it is derived from fissure 3 (Figure 3 and
5). Judging from the contemporary chronicles each of
these tephra falls lasted only for several hours (B4–
5, B29), indicating that the subplinian phases were
short-lived; peaking abruptly and followed by rapid
and steady decline in explosive intensity. Contempo-
rary observations show that in the period June to Au-
gust the eruption columns above the Laki fissures rose
to heights greater than 8.5km (B71). These observa-
tions are compatible with model calculations, which
indicate that the subplinian eruption columns were at
least 11-13 km high and that the principal tephra fall-
outs occurred from high altitude plumes. The vigour
of the explosive activity is verified by the abundance
of Pele’s hair in these tephra falls (e.g., B3–4, B27–
29), because its presence implies very high eruption
velocities in gas-charged fire fountains (Shimozuru,
1994). This evidence, along with extremely effective
vent degassing, is taken to indicate eruption by high
velocity gas-rich jets that drove the subplinian explo-
sive activity (Thordarson et al., 1996).
Again subplinian tephra fall occurred in the Síða
district on 9 July and continued through the 10th
(B48–B50). This tephra fall appears to have had a
southeasterly dispersal because it did not reach the
Meðalland district (Figure 5d). The proximal tephra
stratigraphy shows that at this time all of the fissure
segments southwest of Mt. Laki erupted explosively
and that a significant amount of tephra was dispersed
to the north of the fissures. This dispersal direction is
consistent with the indicated weather pattern for 11–
12 July on reconstructed synoptic weather maps for
the North Atlantic region (Kington, 1988). Thus, the
explosive activity most likely continued through 11
and 12 July.
28 JÖKULL No. 53, 2003

The 1783–1785 Laki-Grímsvötn eruptions
Two more subplinian explosive phases are indi-
cated by the availabledata. There are no records of the
first one in the historic accounts, but a subplinian fall
unit derived from fissure 7 is preserved in the proxi-
mal tephra deposit at the kipuka Innri-Eyrar (Figure
3). The Fissure 7 subplinian fall unit rests directly
on the phreatomagmatic fall unit produced by explo-
sive activity on fissure 6 (see below) and therefore was
formed later, most likely towards the end of August
or at the very beginning of September. The second
phase took place around 7 September as is indicated
by records of tephra fall near Vestmannaeyjar. At the
time the wind was blowing from the northeast (King-
ton, 1988) and the tephra-laden plume would have
been carried directly over the Mýrdalsjökull glacier.
Thus, no or very little tephra would have fallen in
the Fire districts (Figure 5f). This tephra fall unit and
later falls (if they occurred) have not been identified in
the proximal deposit, in part because the three north
easternmost fissures of the Laki vent system are com-
pletely surrounded by lava (Figure 3). Consequently,
the proximal part of the tephra fall that may have been
produced by the explosive activity on these fissures is
either buried beneath the lava or completely removed
by erosion or both.
The explosive activity that commenced on 25 June
and peaked on 27–28 June produced a black erup-
tion plume that “twisted and spun over the moun-
tains forming bolsters” that dispersed “sandy” tephra
over eastern part of the Fire districts (B41–43, Fig-
ure 5c). This description of the plume is consistent
with our knowledge of ash-laden and water-saturated
plumes that are formed by phreatomagmatic erup-
tions (Thordarson, 1991). On the basis of the prox-
imal stratigraphy and timing of events, this tephra fall
is correlated with the phreatomagmatic fall unit that
originated from the tuff cone that is located on the
central part of fissure 4 (TC1 on Figure 3).
Similar plume and tephra-fall was dispersed over
the eastern parts of the Fire Districts on 29 July (B62,
Figure 5e). Phreatomagmatic explosions that pro-
duced the tuff cone on fissure 6 (TC2on Figure 3) pro-
duced this tephra fall. This explosive phase marks a
significant change in the eruption because it coincides
with a shift in the activity at the Lakivents. From this
time on, the main activity was centred on vents to the
northeast of Mt. Laki.
Four of the reported tephra falls are considered to
be from explosive activity at the Grímsvötn volcano.
The tephra fall reported on 18 July is undoubtedly
from Grímsvötn and this is clearly indicated by Stein-
grímsson in his description of the eruption (B52).
The westerly dispersal and the aireal coverage of the
tephra fall on 21 June, 19 July and 14 September
rule out origin at the Laki fissures and point toward
a source farther east (i.e., B34, B54, B70; Figures 5b
and 5d). In light of other evidence presented above
(see p. 18–19) showing that eruptions at the Gríms-
vötn volcano coincided with the activity at the Laki
fissures, it is reasonable to assume that the source of
these tephra falls was explosive activity at Grímsvötn.
The reports of tephra falls clearly show that the ex-
plosive activity at the Laki fissures was episodic and
not simply confined to the beginning of the eruption
as thought previously (Helland, 1886; Thoroddsen,
1879, 1894, 1925; and Thorarinsson, 1967, 1969).
The records appear to give a realistic representation of
periods of vigorous explosive activity at the vents, be-
cause in general terms their occurrence conforms well
to the established tephra stratigraphy and the dispersal
of individual fall units (Thordarson and Self, 1993).
These conclusions are strengthened by the percepti-
ble synchronisation of earthquakes, explosive activity
and increase in lava production (Figure 6, Table 2).
THE LAVA SURGES AND OTHER FLOW
RELATED PHENOMENA
Steingrímsson’s meticulous descriptions of the Laki
lava as it flowed across the plains in front of the Síða
highlands are truly remarkable. Not only do these de-
scriptions provide accurate information about the ad-
vance of the lava but also are a valuable source for
assessing the flow dynamics. A detailed analysis of
these descriptions and their implication for the char-
acteristic emplacement mechanism of the Laki lava is
beyond the scope of this study and will be presented
elsewhere. However, two aspects of Steingrímsson’s
descriptions are particularly important for evaluating
the progress of the Laki eruption; the occurrences and
timing of lava surges and descriptions revealing the
JÖKULL No. 53, 2003 29

Thordarson et al.
1783
17841783
1010 10 10 10
10 10 10 10 1020 20 20 20
20 20 20 20 20
May June July August Sept.
Nov. Dec. Jan. Feb.
Oct.
eruption episode
goshrina
explosive phase
sprengivirkni
subplinian
sprengigos
lava surge
eldhlaup
lava prodiction rate
framleiðni
earthquake swarm
jarðskálftar
low
high
high
low
II III IV VVI VII VIII IX
XXI
?
I
?
?
??
Oct.
phreatomagmatic
þeytigos
Grímsvötn eruption
gos í Grímsvötnum
Figure 6. Schematic illustration of sequence of events during the Laki-Grímsvötn eruptions. Extent of earth-
quake swarms is indicated by wiggly lines; fluctuations in lava discharge shown by shadedarea (not to scale);
eruption clouds denote explosive activity at Laki fissures; eruption clouds with a cone at the base denote ex-
plosive activity at Grímsvötn volcano; arrows indicate onset and termination of Laki eruption. The solid bars
show the extent of each eruption episode, labelled I, II, III etc. Modified from Thordarson and Self (1993). –
Myndræn framsetning á framvindu Skaftárelda.
position of lava flow fronts at various times during
the eruption.
Descriptions of recurring “eldhlaup” (i.e., lava
surges) from the Skaftá River and the Hverfisfljót
River gorges clearly indicate periodic but abrupt in-
crease in lava flow rates. These surges travelled at
velocities in excess of 6 km per day (Thordarson and
Self, 1983) and resulted in rapid advances of the ac-
tive lava fronts that tapered off overseveral days (e.g.,
Appendix C and Figure 5b). It is noteworthy that the
lava surges occurred throughout the eruption and in
each case they emerged from the gorges 3-5 days after
a major explosive phase on the Laki fissures and the
Grímsvötn volcano (Table 2, Figure 6). This synchro-
nisation in the activity strongly indicates that these
lava surges representepisodic increase in lava produc-
tion at the fissures and hence in the magma discharge.
The contemporary accounts do not give unam-
biguous information about the mechanism by which
these surges travelled from the fissures to the active
flow fronts. However, certain inferences can be made
about the most likely transport mode from flow em-
placement structures and the overall volcanic archi-
tecture of the Laki lava flow. The Laki lava flow con-
sists of numerous lava lobes, which range from me-
ters to kilometres in length and decimetre to >20m in
thickness (Thordarson and Self, 1993; Keszthelyi et
al., 2000). Even so, nowhere on the lowlands in front
of the Skaftá River and Hverfisfljót River gorges does
the flow field feature stacked lava flows as defined
by Self et al., (1993), as would be expected if each
of the lava surge had produced their own surface flow
originating at the vents. Furthermore, tumuli, tumulus
ridges, lava rise pits, and lava rise plateaus are com-
mon surface structures in the Laki lava flow field and
filled or drained lava tubes have been found in various
parts of the lava (Figure 7a–e; Thordarson and Self,
1993; Wood et al., 2001). Studies of the flow top
30 JÖKULL No. 53, 2003

The 1783–1785 Laki-Grímsvötn eruptions
3m
10 m
tumulus ridge
b)
lava-rise plateau inflation cleft
inflated flow front
d)
5 m
a)
lava-rise
plateau
lava-rise pit
500 m
c)
tumulus
Figure 7. Flow structures in the Laki lava flow field indicativeof insulated lava transport and flow inflation (i.e.,
endogenous growth). Location of each photograph is shown on Figure 2: (a) tumuli, formed when flow inflation
locally exceeds that of the surrounding lava, for example above a pool in a subsurface lava pathway (Walker,
1991; Hon et al., 1994). (b) A tumulus ridge, which are the surface manifestation of an inflating pathway (e.g.,
Thordarson, 2000). The hornito (arrowed) was formed by a rootless eruption through the central cleft; (c) lava
rise plateaus and lava rise pits (aerial view). The plateaus are formed by uniform inflation of broad pahoehoe
sheet lobes, whereas the pits are localized areas where the rate of inflation was lower than in adjacent parts of
the flow (Walker, 1991), and (d) inflated margin of a lava rise plateau, the cracks are inflation clefts.
JÖKULL No. 53, 2003 31

Thordarson et al.
e)
f)
rp
1 km
internal lava pathway
(filled tube)
1 m
Figure 7. cont. e) An internal lava pathway (i.e., filled tube) in a Laki sheet lobe; f) a large frontal breakout
(large black arrows) near the edge of the Laki flow field (rough textured area, labelled rp). This breakout burst
out from the active flow front on 1 July 1783 and formed a ∼5km long pahoehoe sheet lobe in six days. It is
crowned by numerous smaller breakouts along the lobe margins (the speckle-textured areas). At the southern
margin of the lobe these smaller breakouts are mingledwith breakouts from the main flow field(small arrows
point in direction of flow), indicating that lava was transported through both the main flow field and the sheet
lobe at this time. The main flow field is rubbly pahoehoe covered by thick flow top rubble that was piled up into
concentric ridges oriented perpendicular to flow. The interior of this flow was well insulated and hot because
it fed numerous frontal breakouts of pahoehoe morphology (e.g. large white arrow and small black arrows).
–Flæðistrúktúrar í Skaftáreldahraunisem bera vitni um einangrað flæði undir samfelldri skorpu og hraunris.
32 JÖKULL No. 53, 2003

The 1783–1785 Laki-Grímsvötn eruptions
rubble (Keszthelyi et al., 2000), which covers about
one half of the Laki lava, indicate that it was initially
formed by break up of coherent and stationary pahoe-
hoe crust as a result of surging within the underly-
ing lava. This process exposes the incandescent flow
interior and at this stage the rubble grows in thick-
ness; first by autobrecciation of incandescent lava and
then by the moving lava as it piles the rubble up into
concentric pressure ridges oriented perpendicular to
the flow direction (e.g. Figure 7f). However, the base
of these rubbly lavas is typified by smooth pahoehoe
surface, showing that they have a hybrid morpholog-
ical character that falls between the two end-member
basalt lava flow types, a’a and pahoehoe. However the
emplacement mechanism of these lavas differs signif-
icantly from that of a’a and pahoehoe and therefore it
is considered to be a distinctive flow type, named rub-
bly pahoehoe by Keszthelyi and Thordarson (2000).
All of the flow structures mentioned above are in-
dicative of endogenous growth, suggesting that insu-
lated lava transport and growth by inflation played an
important role during the emplacement of the Laki
lava (e.g. Thordarson and Self, 1993; Keszthelyi et
al., 2000). It is, therefore, reasonable to assume that
the lava, including the surges, was transported from
the vents to the active flow fronts within preferredin-
ternal pathways (i.e., master lava tubes). Such path-
ways were presumably established early on during
emplacement of the lava in the Skaftá River gorge,
and later in the Hverfisfljót River gorge, and subse-
quently lengthened as the lava flow field grew in size.
The lobed architecture of the Laki flow field is also
consistent with this view, because compound flows
are one of the characteristic features of endogenous
flow emplacement (e.g., Walker, 1991; Hon et al.,
1994; Mattox et al., 1993). Each lobe is produced
as a breakout from the main pathways or pre-existing
lobes at the active flow front or onto the lava surface
(Figure 7f), resulting in incremental lengthening and
thickening of the flow field.
A CONCEPTUAL MODEL ON THE PROGRESS
OF THE LAKI ERUPTION
The above analysis of the contemporary accounts
clearly shows that the Laki eruption featured distinc-
tive eruption episodes reflecting periodic increases in
the magma discharge (Figure 6). Each episode be-
gan with an earthquake swarm of increasing inten-
sity that was followed by vigorous explosive activ-
ity at the fissures and sudden increases in outflow of
lava from the fissures. The seismic swarms gener-
ally lasted for several days to a week, with the ex-
ception being the first which lasted for 3–4 weeks.
Each earthquake swarm was followed by a short-lived
subplinian or phreatomagmatic explosive phase lead-
ing into a longer-lasting phase of lava fountaining and
effusive activity. A surge of lava emerged from ei-
ther the Skaftá River or Hverfisfljót River gorge 3–5
days after the beginning of each explosive phase. The
events described above define an eruptive episode and
all in all Laki event featured 10 such episodes (Figure
6). The Laki cone-row is composed of at least 10 en
echelon fissures, trending N47–48◦E (Figure 3). Us-
ing tephra stratigraphy, coupled with descriptions on
the location and timing of explosive activity and lava
surges, each of the ten eruption episodes is linked to
the opening of a new fissure segment andthe stepwise
propagation of the locus of activity from the south-
west to the northeast (Thordarson and Self, 1993). In
this context it is worth noting that there is a percep-
tible increase in the frequency of eruptions at Gríms-
vötn during fall 1783 and that the volcano was active
through to May 1785, adding four eruption episodes
to the 1783–1785 activity on the Grímsvötn volcanic
system (Table 2). It is conceivable that these trends
are linked. The obvious synchronisation in the ac-
tivity at the Laki fissures and the Grímsvötn volcano
indicates that both eruptions resulted from the same
volcano- tectonic event. However, a range of evidence
show that the magma erupted at the Laki fissures is
derived from a large deep-seated reservoir located at
the crust-mantle boundary, whereas the eruptions at
Grímsvötn occurred from a shallow crustal magma
chamber (Gudmundsson, 1987; Sigmarsson et al.,
1991; Thordarson and Self, 1993). Thus, the trends
mentioned above may indicate a gradual restoration in
the flow of magma from the deep-seated reservoir to
the shallow magma chamber beneath the Grímsvötn
volcano.
JÖKULL No. 53, 2003 33

Thordarson et al.
CONCLUDING REMARKS
This summary on the course of events during the
1783–1785 Laki-Grímsvötn eruptions is primarily
based on an analysis of the information contained in
the contemporary accounts, although some of the con-
clusions are reinforced by field observations. The
1783–1784 Laki eruption is a classic example of
basaltic flood lava eruptions and the only one that was
observed and described by man. Consequently, the
Laki accounts provide us with unprecedented and vi-
tal insight into the nature and progressionof flood lava
eruptions. This data has been applied successfully in
studies of past flood lava eruptions, not only here on
Earth but also on the other planets (e.g. Thordarson
and Self, 1998; Keszthelyi et al., 2000).
Our work has produced results that differ from
those put forth previously by other geologists (e.g.
Thoroddsen, 1879, 1925; Helland, 1886; Thorarins-
son, 1967; 1969). This difference can in parts be ex-
plained by our approach, because for the first time
the information contained in the written records have
been systematically evaluated according to the pro-
cesses they describe. This has allowed us to present
a detailed reconstruction of the Laki eruption and as-
sociated activity at the Grímsvötn volcano. In the
last thirty years our understanding of volcanological
processes has improved significantly. Consequently
many of the original Laki descriptions have gained
a new perspective and been re-interpreted in light of
latest advances. One purpose of this work is to pro-
vide the volcanological communitywith the informa-
tion contained in the original accounts on the progress
of the 1783–1785 Laki-Grímsvötn eruptions such that
those who are willing can themselves evaluate our re-
sults and conclusions. Furthermore, we hope that this
contribution will provoke further studies on this re-
markable event.
ACKNOWLEDGEMENTS
This contribution is in part derived from work con-
ducted by the senior author for his B.Sc. Honors and
M.Sc. degrees and we want to thank the University
of Iceland and the University of Texas at Arlington
for their support during the course of these studies.
We also thank Leó Kristjánsson and Olgeir Sigmars-
son for constructive reviews and the editors Áslaug
Geirsdóttir and Bryndís Brandsdóttir for their assis-
tance and patience.
ÁGRIP
Skaftáreldar 1783–1784, ásamt þeytigosum í Gríms-
vötnum voru afleiðing tveggja ára umbrotahrinu í
Grímsvatnakerfinu frá miðjum maí 1783 til 26. maí
1785. Skaftáreldagosið er, á eftir Eldgjárgosinu 934–
940, annað stærsta flæðigosið á sögulegum tíma og
mestu náttúruhamfarir sem sögur fara af á Íslandi.
Samtímaheimildir veita nákvæmar upplýsingar um
atburðarásina í þessum umbrotum, sérstaklega hvað
varðar framvindu Skaftárelda. Við höfum gert ítarlega
úttekt á öllum aðgengilegum frásögnum sem varpa
ljósi á atburðarás og eðli þessara umbrota. Til þess að
fá heildaryfirlit yfir þessar upplýsingar voru frásagn-
irnar skráðar í tímaröð og þeim raðað niður í flokka
eftir þeim atburðum sem fjallað er um í heimildunum,
þ.e. skjálftavirkni, gjóskugosvirkni, hraunstreymi og
áttaskyni (sjá viðauka, töflur A til D). Þessar upplýs-
ingar voru notaðar til þess að kryfja til mergjar ein-
staka atburði og gosferli í Skaftáreldum. Helstu nið-
urstöður athugana okkar sýna að umbrotin á Gríms-
vatnakerfinu gengu á með hrinum. Samtímaheimildir
greina frá að minnsta kosti 14 goshrinum. Tíu af þess-
um goshrinum voru sjálfir Skaftáreldarnir, þar sem
sérhver hrina hófst með skjálftum og öflugu þeytigosi.
Í kjölfar hverrar hrinu jókst hraunrennslið verulegafrá
gosstöðvunum niður á láglendið. Grímsvötn gusu að
minnsta kosti sex sinnum á meðan Skaftáreldar stóðu
yfir og tvívegis eftir að eldar kulnuðu á gossprung-
unum inni á Síðumannaafrétti (þ.e. apríl 1784 og maí
1785). Í fjórum tilvikum var þeytigos í Grímsvötnum
samfara hrinu á Skaftáreldasprungunni og telst því
vera hluti af þeim goshrinum. Í hinum fjórum tilfell-
unum takmarkaðist aukin gosvirkni aðeins við Gríms-
vötn og af þeim sökum eru þessi gos talin sem sérstak-
ar goshrinur.
REFERENCES
Bjornsson, A., K. Saemundsson, P. Einarsson, E. Tryggva-
son and K. Gronvold 1977. Current rifting episode in
north Iceland. Nature 266, 318-323.
Björnsson, E. 1783. Relation eins prests sem ár 1783
ferðaðist um sumarið á Suðurland frá Múlasýslu yfir
34 JÖKULL No. 53, 2003

The 1783–1785 Laki-Grímsvötn eruptions
Skaftafells bæði áfram og heimleiðis (Relation by one
pastor, who travelled in the summer 1783 in Southern
Iceland from Múlashire through Skaftafellshire back
and forth). In T. Einarsson, G. M. Gudbergsson, G.
Á. Gunnlaugsson, S. Rafnsson and S. Thorarinsson
(eds.), Skaftáreldar 1783–1784, Ritgerðir og Heimild-
ir. Mál og Menning, Reykjavík 1984, 295–297.
Brandsdóttir, B. 1992. Historical accounts of earthquakes
associated with eruptive activity in the Askja volcanic
system. Jökull 42, 1–12.
Brandsdóttir, B. and P. Einarsson 1979. Seismic activity
associated with the september 1977 deflation of the
Krafla central volcano in North-eastern Iceland. J. Vol-
canol. Geotherm. Res. 6, 197–212.
Christ, J. L. 1783. Von der ausserordentlichen Witt-
terung des Jahres 1783, in Anstehnung des anhal-
tenden und heftigen Höerauchs; von Thermometer und
Barometer, von dem naturlichen Barometer unserer
Geend, dem Feldberg oder der Höhe, und vin die
Beschefferheit und Entstehung unser ungewöhnlichen
Lufterscheiningungen, wie auch etwas von dem Erd-
beben. Heimannischen Buchhandlung, Frankfurt and
Leipzig, 72 pp.
Einarsson, T., G. M. Guðbergsson, G. Á. Gunnlaugsson, S.
Rafnsson, and S. Þórarinsson (eds.) 1984. Skaftáreld-
ar 1783–1784, Ritgerðir og Heimildir. Mál og Menn-
ing, Reykjavík, 442 pp.
Eiríksson, J. 1783a. Efterretning um Ilds-Udbrydelsen i
Vester Skaptefells-Syssel i Island – som efter Konglige
Allernaadigste Befalning skal oplæses i alle Kirker i
Kiøbenhavn paa første Søndag efter Nyt-Aar 1784.
(A narrative on the eruption in Western-Skaftafellshire
in Iceland – which in accordance to His Royal High-
ness compassionate order shall be read in all churches
in Copenhagen on the first Sunday of the New Year
1784). B. H. Haecke, Copenhagen, 8 pp.
Eiríksson, J. 1783b. Extract of Forpagter Erichsens Skriv-
else dat. Fliotestad i Vester-Skaptefells-syssel 11 Jul.
1783. (An extract from Eiríksson’s description (of
the Laki eruption) dated at Ljótarstaðir in Western
Skaftafellshire 11 July 1783). In T. Einarsson, G.
M. Gudbergsson, G. Á. Gunnlaugsson, S. Rafnsson
and S. Thorarinsson (eds.), Skaftáreldar 1783–1784,
Ritgerðir og Heimildir, Mál og Menning, Reykjavík
1984, 274-276.
Fiacco, J. J., T. Thordarson, M. S. Germani, S. Self, J.
M. Palais, S. Withlow and P. M. Grootes 1994. Atmo-
spheric aerosol loading and transport due to the 1783–
1784 Laki eruption in Iceland, interpreted from ash
particles and acidity in the GISP2 ice core. Quat. Res.
42, 231–240.
Finnsson, H. 1796. Um mannfækkun í hallærum (Decima-
tion of the population in Iceland due to famines). Rit
þess konunglega íslenska lærdóms-lista félags XIV,
30–226.
Franklin, B. 1784. Meteorological imagination’s and con-
jectures. Manchester Literary and Philosophical Soci-
ety Memoirs and Proceedings 2, 122.
Grattan, J. P. 1998. The distal impact of volcanic gases
and aerosols in Europe: a review of the 1783 Laki fis-
sure eruption and environmental vulnerability in the
late 20th century. Geol. Soc. Spec. Publ. 15, 7–53.
Grattan, J. P. and M. B. Brayshay 1995. An amazing
and portentous summer: Environmental and social re-
sponses in Britain to the 1783 eruption of an Iceland
Volcano. The Geograph. J. 161, 125–134.
Gudmundsson, A. 1987. Lateral magma flow, caldera col-
lapse, and a mechanism of large eruptions in Iceland.
J. Volcanol. Geotherm. Res. 34, 65–78.
Gudmundsson, A. 2000. Dynamics of volcanic systems in
Iceland: Example of tectonism and volcanism at jux-
taposed hot spot and mid-ocean ridge systems. Ann.
Rev. Earth Planet. Sci. 28, 107–140.
Guðmundsson, L. 1783. Reports by Lýður Guðmundsson
sheriff of West-Skaftafellshire, dated 2 August and 1
September, 1783. In T. Einarsson, G. M. Guðbergsson,
G. Á. Gunnlaugsson, S. Rafnsson, and S. Þórarins-
son (Editors), Skaftáreldar 1783–1784, Ritgerðir og
Heimildir, Mál og Menning, Reykjavík, 1984, 283–
287.
Gudmundsson M. T., F. Sigmundsson and H. Bjornsson
1997. Ice-volcano interaction of the Gjálp subglacial
eruption, Vatnajökull, Iceland. Nature 389, 954–957.
Gunnlaugsson, G. Á. and S. Rafnsson 1984. Heimildir
til sögu Skaftárelda og Móðuharðinda 1783–1785
(Contemporary sources to the history of the Skaftár
Fires and the Haze Famine). In T. Einarsson, G. M.
Guðbergsson, G. A. Gunnlaugsson, S. Rafnsson, and
S. Þórarinsson (Editors), Skaftáreldar 1783–1784, Rit-
gerðir og Heimildir, Mál og Menning, Reykjavík,
263–442.
Helland, A. 1886. Lakis kratere og lavaströmme. (The
Laki cones and lava flows). Universitets program,
Kristiania, Norway, 40 pp.
Hon, K., J. Kauahikaua, R. Denlinger and K. Mackay
1994. Emplacement and inflation of pahoehoe sheet
flows: Observations and measurements of active lava
JÖKULL No. 53, 2003 35

Thordarson et al.
flows on Kilauea volcano, Hawaii. Geol. Soc. Am.
Bull. 106, 351–370.
Hólm, S. M. 1784. Om Jordbranden paa Island i Aaret
1783. (About the Earth Fire in Iceland in the year
1783). Peder Horrebow, Copenhagen, 84 pp.
Jakobsson, S. P. 1979. Petrology of recent basalts of the
Eastern Volcanic Zone, Iceland. Acta Naturalia Is-
landica 26, 103pp.
Jóhannesson, H. 1984. Grímsvatnagos 1933 og fleira
frá því ári (Grímsvötn eruption and associated phe-
nomen). Jökull 34, 151–158.
Keszthelyi, L. and T. Thordarson 2000. Rubbly pahoehoe:
A previously undescribed but widespread lava type
transitional between a’a and pahoehoe. GSA Annual
Meeting abstract volume, November 15 2000.
Keszthelyi, L., McEwen, A. S. and T. Thordarson 2000.
Terrestrial analogs and thermal models for Martian
flood lavas. J. Geophys. Res. 105, 15027–15049.
Kington, J. A. 1988. The weather of the 1780’s over Eu-
rope. Cambridge University Press, Cambridge, Eng-
land.
Larsen, G., M. T. Gudmundsson and H. Bjornsson 1998.
Eight centuries of periodic volcanism at the center of
the Iceland hotspot revealed by glacier tephrostrati-
graphy. Geology 26, 943-946.
Larsen, G., E. G. Vilmundardottir and B. Thorkelsson
1992. Heklugosið 1991: Gjóskufallið og gjóskulagið
frá fyrsta degi gossins (The Hekla eruption of 1991:
The tephra fall). Náttúrufræðingurinn 61, 159–176.
Lievog, R. 1784. Meteorological Observations in South-
ern Iceland. Weather logs preserved at the National
Library of Iceland (Lbs I.B. 234 A, 4to).
Mattox, T. N., C. Heliker, J. Kauahikaua and K. Hon 1993.
Development of the 1990 Kalapana flow field, Kilauea
volcano, Hawaii. Bull. Volcanol. 55, 407–413.
Montredon, M. d. 1784. Recherches sur l’origine & sur la
nature des Vapeurs qui ont régné dans l’Atmosphère
pendant l’été de 1783. Mémoires de l’Académie
Royale des Sciences MDCCLXXXI, 754–773.
Pálsson, S. 1784. Historia ignis in oriente Islandiæ erump-
entis Anno 1783, qvoad innotuit in tractu Skagafjörd-
ensi, complectitur in se et effectus varios (The story
of the earth fire which broke out in Eastern Iceland
in the year 1783, as long as it was observed in Skaga-
fjörður; concerning the progress of the eruption and its
various effects). In T. Einarsson, G. M. Gudbergsson,
G. Á. Gunnlaugsson, S. Rafnsson and S. Thorarinsson
(eds.), Skaftáreldar 1783–1784, Ritgerðir og Heim-
ildir, Mál og Menning, Reykjavík 1984, 419–422.
Pálsson, S. 1794. Ferðabók Sveins Pálssonar, Dagbækur
og Ritgerðir 1791–1797. (The Travelogue of Sveinn
Pálsson. Notebooks and essays through 1791–1797).
2nd edition, Örn og Örlygur,Reykjavík 1983, 813 pp.
(First published by Snælandsútgáfan in 1945.)
Pétursson, M. 1730–1784. Höskuldsstaðarannáll 1730–
1784. Annálar 1400–1800, (Annales Islandici) IV.
Hið Íslenska Bókmenntafélag, Reykjavík 1940-1948,
463–603.
Rafnsson, S. 1984. Um Eldritin 1783–1784 (A review
of the Eldrits from 1783–1784). In T. Einarsson, G.
M. Guðbergsson, G. A. Gunnlaugsson, S. Rafnsson
and S. Þórarinsson (Editors), Skaftáreldar 1783–1784.
Ritgerðir og Heimildir, Mál og Menning, Reykjavík,
243–262.
Self, S., L. Keszthelyi and T. Thordarson 1998. The im-
portance of pahoehoe. Ann. Rev. Earth Planet. Sci. 26,
81–110.
Shimozuru, D. 1994. Physical parameters governing the
formation of Pele’s hair and tears. Bull. Volcanol. 56,
217–219.
Sigmarsson, O., M. Condomines, K. Gronvold and T.
Thordarson 1991. Extreme magma homogeneity in the
1783–1784 Lakagígar eruption: Origin of a large vol-
ume of evolved basalt in Iceland. Geophys. Res. Lett.
18, 2229–2232.
Sigmarsson, O., H. R. Karlsson and G. Larsen 2000. The
1996 and 1998 subglacial eruptions beneath the Vatna-
jökull ice sheet in Iceland: contrasting geochemical
and geophysical inferences on magma migration. Bull.
Volcanol. 61, 468–476.
Sigurdsson, H. 1982. Volcanic pollution and climate: The
1783 Laki eruption. Eos Transactions AGU 63, 601–
602.
Stefánsson, R. and P. Halldórsson 1988. Strain release
and strain-build up in the south Iceland seismic zone.
Tectonophysics 152, 267–276.
Steingrímsson, J. 1757. Um Kötlugjá (About the Katla Fis-
sure), Safn til Sögu Íslands IV, Copenhagen, 1907–
1915, 190–199.
Steingrímsson, J. 1783a. Lítið ágrip um nýja eldsuppkomu
í vestariparti Skaftafellssýslu og þess verkanir sem
framkomnar eru (A short compendium of the recent
volcanic outburst in western part of Skaftafellshire). In
T. Einarsson, G. M. Gudbergsson, G. Á. Gunnlaugs-
son, S. Rafnsson and S. Thorarinsson (eds.), Skaft-
áreldar 1783–1784, Ritgerðir og Heimildir, Mál og
Menning, Reykjavík 1984, 272–274.
36 JÖKULL No. 53, 2003

The 1783–1785 Laki-Grímsvötn eruptions
Steingrímsson, J. 1783b. Póstur úr bréfi til séra Bjarna
Jónssonar (Abstract from a letter to Reverend Bjarni
Jónsson). Safn til Sögu Íslands IV, Copenhagen 1907–
1915, 69–79.
Steingrímsson, J. 1788. Fullkomið skrif um Síðueld (A
complete description on the Síða volcanic fire). Safn
til Sögu Íslands IV, Copenhagen 1907–1915, 1–57.
Steingrímsson, J. 1988. Fires of the Earth: the Laki erup-
tion 1783–1784. Nordic Volcanological Institute and
the University of Iceland Press. (English translation of
the original 1788 text).
Steingrímsson, J. and S. Ólafsson 1783. Einföld og sönn
frásaga um jarðeldshlaupið í Skaftafellssýslu árið
1783. (A simple, but true narrative of the eruption in
Skaftafellshire in the year 1783). Safn til Sögu Íslands
IV, Copenhagen 1907–1915, 58–69.
Steinthorsson, S. 1992. Annus Mirabilis: 1783 í erlendum
heimildum (Annus Mirabilis: the year 1783 according
to contemporary accounts outside of Iceland). Skírnir
166, 133–159.
Stephensen, M. 1785. Kort beskrifelse over den nye vul-
cans Ildsprudelse i Vester-Skaptefields-Syssel paa Is-
land i Aaret 1783. (A short description of the new Vol-
canic eruption in Western Skaftafellshire in the year
1783). Nicolaus Möller, Copenhagen, 148 pp.
Stephensen, M. 1786. Betænking over Indholdet af
Chirurgi Jon Petersens Afhandling om Ildrögens kien-
deligste Virkninger af Ildsprudningen 1783 saavidt
Norder-Island angaaer (Comments on the contents of
Dr. Jón Pétursson’s essay on the effects of the volcanic
haze from volcanic eruption in 1783 in North Iceland).
Islands Journal 7, Þjóðskjalasafn Rtk. 42.2, fylgiskjöl
nr. 104.
Stephensen, O. 1783. Abstract from prefect Stephensen
letter to Erichsen, the deputy of the treasury, dated
15 August 1783. In T. Einarsson, G. M. Gudbergs-
son, G. Á. Gunnlaugsson, S. Rafnsson and S. Thor-
arinsson (eds.), Skaftáreldar 1783–1784, Ritgerðir og
Heimildir, Mál og Menning, Reykjavík 1984, 279.
Sünckenberg, J. C. 1783. Abstract from a letter by
J. C. Sünckenberg to the directors of the Royal
Monopoly Company in Iceland, dated in Stykkis-
hólmur, W-Iceland 24 July 1783. In T. Einarsson, G.
M. Guðbergsson, G. A. Gunnlaugsson, S. Rafnsson,
and S. Þórarinsson (Editors), Skaftáreldar 1783–1784,
Ritgerðir og Heimildir, Mál og Menning, Reykjavík
1984, 269–270.
Svendborg 1783. Abstract from Svendborg’s letter to agent
Pontoppidan, dated in Hafnarfjörður, Iceland, 31. Au-
gust 1783. In T. Einarsson, G. M. Guðbergsson, G. A.
Gunnlaugsson, S. Rafnsson, and S. Þórarinsson (Edi-
tors), Skaftáreldar 1783–1784, Ritgerðir og Heimildir,
Mál og Menning, Reykjavík 1984, 279-280.
Sæmundsson, J. 1729. Ein ný relation um brennisteins-
fjöllin við Mývatn á Íslandi og Norðursýslu (A new
description on the sulfur mountains at Lake Mývatn
in Northern Iceland). Safn til Sögu Íslands IV, Copen-
hagen 1907–1915, 389–398.
Svendborg 1783. Abstract from Svendborg’s letter to agent
Pontoppidan, dated in Hafnarfjörður, Iceland, 31. Au-
gust 1783. In T. Einarsson, G. M. Guðbergsson, G. A.
Gunnlaugsson, S. Rafnsson, and S. Þórarinsson (Edi-
tors), Skaftáreldar 1783–1784, Ritgerðir og Heimildir,
Mál og Menning, Reykjavík 1984, 279–280.
Thorarinsson, Á. 1783. Hr. Provst Torarensens Beretning
(Reverend Thorarensens Documentation). In T. Ein-
arsson, G. M. Gudbergsson, G. Á. Gunnlaugsson,
S. Rafnsson and S. Thorarinsson (eds.), Skaftáreldar
1783–1784, Ritgerðir og Heimildir, Mál og Menning,
Reykjavík 1984, 280–283.
Thorarinsson, S. 1955. Öskufall svo sporrækt varð og
Kötlugosið 1721 (Ash-fall so footprints were traceable
and the 1721 Katla eruption). Náttúrufræðingurinn 25,
87–98.
Thorarinsson, S. 1967. Skaftáreldar og Lakagígar: Myndir
úr jarðfræði Íslands VIII (The Skaftár Fires and Laki
cone-row: Illustrations from the geology of Iceland
VIII). Náttúrufræðingurinn 37, 1–57.
Thorarinsson S. 1969. The Lakagígar eruption of 1783.
Bull. Volcanol. 3, 910–929.
Thorarinsson, S. 1974. Vötnin Stríð: Saga Skeiðarár-
hlaupa og Grímsvatnagosa. (The history of Grímsvötn
eruptions and jökulhlaup’s). Bókaútgáfa Menningar-
sjóðs, Reykjavík, 254 pp.
Thorarinsson, S. 1976. Eruption of Hekla 1947-1948:
Course of events. Vísindafélag Íslendinga, IV, 1, 5–31.
Thorarinsson, S. 1979. On the damage caused by volcanic
eruptions with special reference to tephra and gases. In
P. D. Sheets and D. K. Grayson (eds.), Volcanic Activ-
ity and Human Ecology, Academic Press, New York,
125–159.
Thorarinsson, S., 1984. Annáll Skaftárelda (The annals of
the Skaftá Fires). In T. Einarsson, G. M. Gudbergs-
son, G. Á. Gunnlaugsson, S. Rafnsson and S. Thor-
arinsson (eds.), Skaftáreldar 1783–1784, Ritgerðir og
Heimildir, Mál og Menning, Reykjavík 1984, 11–36.
Thorarinsson, S. and K. Sæmundsson 1979. Volcanic ac-
tivity in historical times. Jökull 29, 29–32.
JÖKULL No. 53, 2003 37

Thordarson et al.
Thordarson, T. 1990. The eruption sequence and eruption
behaviour of the Skaftá Fires, 1783–1785, Iceland:
Characteristics and distribution of eruption products.
M.Sc. Thesis, The University of Texas at Arlington,
Arlington, 239 pp.
Thordarson, T. 1991. Gjóskan og framvinda gossins (The
1783–1784 Laki eruption: Tephra production and
course of events). University of Iceland Geoscience
Special Publication, F90018. University of Iceland,
Reykjavík, 187 pp.
Thordarson, T. and S. Self 1993. The Laki (Skaftár Fires)
and Grímsvötn eruptions in 1783–1785. Bull. Vol-
canol. 55, 233–263.
Thordarson, T. 1995. Volatile release and atmospheric ef-
fects of basaltic fissure eruptions. Ph.D. Thesis, Uni-
versity of Hawaii, Honolulu, 570 pp.
Thordarson, T. 2000. Physical volcanology of lava flows
on Surtsey, Iceland, Surtsey Research Progress Report
XI, 125–142.
Thordarson, T. and S. Self 1988. Old maps help interpret
Icelandic eruptions. Eos Transactions AGU 69, 86.
Thordarson, T. and S. Self 1993. The Laki (Skaftár Fires)
and Grímsvötn eruptions in 1783–1785. Bull. Vol-
canol. 55, 233–263.
Thordarson, T. and S. Self 1998. The Roza Member,
Columbia River Basalt Group: A gigantic pahoehoe
lava flow field formed by endogenous processes? J.
Geophys. Res. 103, 27,411–27,445.
Thordarson, T. and S. Self 2001. Real-time observations of
the Laki sulfuric aerosol cloud in Europe 1783 as doc-
umented by Professor S. P. van Swinden at Franeker,
Holland. Jökull 50, 65–72.
Thordarson, T. and S. Self 2003. Atmospheric and envi-
ronmental effects of the 1783–1784 Laki eruption: a
review and reassessment. J. Geophys. Res., 108 (D1)
10.1029/2001JD002042.
Thordarson Th., S. Self and S. Steinthorsson 1993.
Aerosol loading of the Laki fissure eruption and its
impact on climate. Eos Transactions, AGU 74, 106
Thordarson, T., S. Self, N. Oskarsson and T. Hulse-
bosch 1996. Sulfur, chlorine, and fluorine degassing
and atmospheric loading by the 1783–1784 A.D. Laki
(Skaftár Fires) eruption in Iceland. Bull. Volcanol. 58,
205–225.
Thordarson, T., D. J. Miller and G. Larsen 1998. New data
on the age and origin of the Leidolfsfell cone group in
south Iceland. Jökull 46, 3–15.
Thordarson, T., D. J. Miller, G. Larsen, S. Self and H.
Sigurdsson 2001. New estimates of sulfur degassing
and atmospheric mass-loading by the A.D. 934 Eld-
gjá eruption, Iceland. J. Volcanol. Geotherm.Res. 108,
33–54.
Thoroddsen, T. 1879. De vulkanske Udbrud paa Island i
Aaret 1783 (Volcanic eruptions in Iceland in the year
1783). Geografisk Tidskrift 3, 67–89.
Thoroddsen, T. 1894. Ferð um Vestur Skaftafellssýslu
sumarið 1893 (Travelogue from Western Skaftafell-
shire in the summer 1893). Andvari 19, 44–161.
Thoroddsen, T. 1908–1911. Lýsing Íslands (Physiographic
description of Iceland). Hid Íslenska Bókmenntafélag,
Copenhagen, 1038 pp.
Thoroddsen, T. 1914. Eldreykjarmóðan 1783. (The vol-
canic haze, 1783). In Afmælisrit til Dr. phil K. Kaa-
lund, Hid íslenska fræðafélag, Copenhagen, 88-107.
Thoroddsen, T. 1925. Die Geschichte der Isländischen
Vulkane. Kongelige Danske Videnskabernes Selskabs
Skrifter, Ottende Række, Niende Bind. A. F. Høst and
Søn Kongelige Hof-Boghandel, Copenhagen, 458 pp.
Thorsteinsson, P. 1783. Ketilsstaðaannáll 1742–1784. An-
nálar 1400–1800 (Annales Islandici), IV. Hið Íslenska
Bókmenntafélag, Reykjavík 1940–1948, 389–462.
Traumüller, F. 1885. Die trockenen Nebel, Dämmerungen
und vulkanische Ausbrüche des Jahres 1783. Meteo-
rologische Zeitschrift 2, 138–140.
Walker, G. P. L. 1991. Structure and origin by injection of
lava under surface crust, of tumuli, “lava rises”, “lava-
rise pits”, and “lava-inflation clefts” in Hawaii. Bull.
Volcanol. 53, 546–558.
Wood, C. A. 1992. Climatic effects of the 1783 Laki erup-
tion. In C. R. Harington (Editor), The year without a
summer? Canadian Museum of Nature, Ottawa, 58–
77.
Wood, C., P. Cheetham, R. Watts and N. Randall 2001.
Laki underground 2000: The Bournemouth/Dundee
Universities joint expedition to Iceland. Expedition
Report ISBN: 1-85899-127-7, School of Conservation
Sciences, Bournemouth University, 36 pp.
38 JÖKULL No. 53, 2003

The 1783–1785 Laki-Grímsvötn eruptions
APPENDIX
The appendix contains English translations (by the senior author) of original text in the contemporary accounts containing
information on seismicity (Table A), explosive activity (Table B), lava flows (Table C), and sense of directions in the Fire
districts (Table D). The Index number links to citation numbers in the text. The first number in the Source column refers
to the numbered citations listed below and the latter indicates the page number in the original source for each quotation: 1.
Steingrímsson, 1783a; 2. Steingrímsson and Ólafsson, 1783; 3. Steingrímsson, 1783b; 4. Steingrímsson, 1788; 5. Eiríksson,
1783b; 6. Björnsson, 1783; 7. Thorarinsson, 1783; 8. Pálsson, 1783; 9. Hólm, 1784; 10. Stephensen, 1785; 11. Pálsson,
1794; 12. Stephensen, 1783; 13. Pétursson, 1783; 14. Thorsteinsson, 1783. Full reference to the original sources is given
in the bibliography in the back. Note that the term Vesturgjá is a synonym for Útnorðursgjá and similarily Austurgjá is a
synonym for Landnorðursgjá. Table A: Seismic activity
Date Index Source Quotation
19 April–29
May A1 5, p. 274 In mid-May I [Jón Eiríksson at Ljótarstaðir in Skaftártunga] felt some earthquakes, during nights especially. I kept this to
myself and did not hear any others speak of it until Ascension Day when some earthquakes were felt in the upper part of
Skaftártunga. From then on earthquakes increased (in frequency and intensity) so people slept in tents out on open fields.
Early June A2 13, p. 455 ...but the farm Ásar in Skaftártunga almost collapsed from the intensive earthquakes, which were associated with this
eruption.
1 June A3 2, p. 58 This day was the beginning of our afflictions, when strong earthquakes were felt, not only here in Síða, Landbrot, Meðal-
land and Álftaver, but also in Mýrdalur in the west and Öræfi in the east. These earthquakes increased in frequency and
intensity) the following week and were mostly felt in the mornings and the evenings.
9 June A4 4, p. 9 The night before strong earthquakes and tremors were felt.
9–11 June A5 5, p. 274 The earthquakes increased steadily during the first four days of the eruption. [until 11 June]
10 June A6 2, p. 58 ...earthquakes continued to be felt... [i.e., in the Síða district]
11 June A7 2, p. 59 All that day and the following night strong earthquakes were felt.
13 June A8 4, p. 9 ...cracking and thuds were heard behind the mountains in the northwest followed by earthquakes [i.e., associated with
activity at Útnorðursgjá]
14 June A9 2, p. 60 The lava emerged out from the Skaftá River gorge and along Árfjall mountain with such a murmur that the earth trembled
and vibrated along with continuous sparks of fire and claps of thunder, sofew people rested or slept that night.
15 June A10 4, p. 11 All that day and the following night earthquakes were felt.
16 June A11 4, p. 12 The din [tremor?] and the cracking, the fires and the stench along with the earthquakes north of the mountains bordering
Síða was so frenzied, that no one was sure whether this settlement was safe.
18 June A12 4, p. 13 ...again the lava surged out of the western fissure, filling up the Skaftá River gorge which was both deep and wide, and
accompanied by intense earthquakes, tremors and frequent thunder claps. ...This day and night the thunderclaps were so
extensive that everything trembled and vibrated and earthquakes caused cracking noise from every house timber.
19 June A13 2, p. 62 Weather was calm, but earthquakes were frequent and hard.
23–26 June A14 2, p. 63 ...lava emerged from the Skaftá River gorge at steady rate accompanied by earthquakes and thunder and lightning were
observed all around.
30 June A15 4, p. 19 ...calm weather, thunderclaps and earthquakes along with a boiling noise from the fissure.
14 July A16 4, p. 21 ...for the fourth and last time a lava surge emerged out from Vesturgjá along with boiling, creaking and cracking noises and
such a din that everything appeared to be going out of control, [din = tremor?].
17 July A17 4, p. 22 What was left of the livestock crowded together out of their dismay or ran around shrieking in their own sound and some
dropped dead where they stood. [old Icelandic accounts show that when animals gathered in groups, it is indicative of
seismic activity]
18 July A18 4, p. 22 The din [tremor?] was so frenzied that I [Jón Steingrímsson] thought everything would be demolished.
20 July A19 4, p. 22 ...continuous earthquakes.
1–3 Aug. A20 4, p. 26 ...the same boiling was continuously heard from this gjá along with earthquakes, thunder, lightning and continuous out-
pouring of lava [gjá = Landnorðursgjá].
23–31 Aug. A21 4, p. 28 ...the eruption proceeded in similar fashion, except intensive earthquakes were felt and claps were heard from Austurgjá.
26 Sept. A22 4, p. 29 ...intensive tremors or rather earthquakes were felt again, especially in the eastern part of Síða and in Fljótshverfi, which
was followed by increasing activity and lava effusion to the north of the mountains...
24 Oct. A23 4, p. 30 This day earthquakes started again and strong stench was smelled from east-northeast...
Nov. A24 4, p. 31 ...occasional weak earthquakes were felt in Síða.
24 Nov. A25 4, p. 31 A strong earthquake was felt in Meðalland, but its intensity was much less here at the foot of the scarp.
Jan.–Feb.
1784 A26 4, p. 32 After a long period of freezing and heavy winds that lasted through January, strong earthquakes were felt again, though dif-
ferent than previous ones. Some moved the earth slightly up and down, others travelled like oceanic waves from northwest
to southeast. The frozen ground cracked with sound of thumps and claps.
14–25 Aug.
1784 A27 4, p. 35 In the summer 1784, 14 and 15 August weak earthquakes were felt here in Síða, which were related to activity at the
north-northeastern fissure, and the last one occurred on the 25 August. After that such activity [i.e., earthquakes] was not
perceived until it recurred in the Rangárvallasýsla and Árnessýsla were it ruined many farmhouses.
JÖKULL No. 53, 2003 39

Thordarson et al.
Table B: Explosive activity
Date Index Source Quotation
spring
1783 B1 2, p. 58 During this time people [in the Fire districts] often noticed a bluish smoke over the ground, and many thought, as became
evident later on, that it was from a volcanic eruption. Later that spring travellers from Reykjanes told us that the ongoing
submarine eruption at Helgusker southwest of Reykjanes was not the only one. The crew on a brig coming from Denmark
saw a column of fire in the mountains or the glaciers north of us [i.e., the Fire districts], as they passed the southern coast
on their way to Hafnarfjörður. [Thoroddsen (1908–1911) states that the submarine eruption off Reykjanes was seen in early
May 1783 by a crew of a brig that was on its way to Hafnarfjörður in southwestern Iceland. Steingrímsson’s description
implies that it was the same crew that saw the columns of fire above the Fire districts, suggesting an eruption in the area in
early May.]
spring
1783 B2 10, p. 11 ...the water level was unusually high in the Skaftá River and the water dirty and ill-smelling.
8 June B3 1, p. 272 After a period of intensive earthquakes, on 8 June an eldgjá began erupting in the so-called Síða highlands due southwest
of Úlfarsdalur glacier. Volcanicash and hairy material [Pele’s hair] was dispersed over this district [i.e., Síða] so footprints
marked in it on the third day of the eruption. The heavy rain which fell from the eruption from the fire and smoke columns
that rose from the fissure, which now exposed itself by numerous columns of fire, contained salty and sulfur smelling water
which caused smarting in the eyes and on the skin.
8 June B4 2, p. 58 ...about 9 a.m. a black plume emerged from north, dispersed over Síða so it became dark inside our homes. So much
volcanic ash fell out of the plume, that footprints marked in it. The ash resembled the remains of burned coal and contained
glittering hairs that were easily crushed between the fingers. Such an ash fell here many times after that. This plume was
directed away later that day by southerly breeze.
8 June B5 4, p. 8 ...about 9 a.m. a black sand-bearing mist and a plume emerged north of the highest mountains bordering the Síða district,
so big that in a short time it spread over Síða and part of Fljótshverfi. The cloud caused it to be dark inside our homes and so
much volcanic ash fell out of the cloud that footprints marked in the ash deposit. A drizzle that fell from the plume on this
day in Skaftártunga was so laden with ash that it resembled ink. Later that day the plume was kept away by southeasterly
breeze.
8 June B6 3, p. 69 ...a dense ashy plume fell over us, so it became dark inside houses and footprints marked in the deposit. For the nextthree
days the ash-fall was accompanied by heavy rainfall; and occasionally a strong stench, like in a steamy sulfur mine, was
felt. The rainwater was salty and bitter and the drops felt like hot brine when they fell on bare skin.
8 June B7 5, p. 274 On Whitsunday I first saw a fire column and smoke northeast from my home [Ljótarstaðir in Skaftártunga], but north of
Síða, which were thought to be located within the glacier where Grímsvötn are to be found. ...No ash was observed to fall
here [in Skaftártunga], but in Síða the ash deposit was so thick that footprints marked in it when someone walked over it.
The ash was washed into the ground by the rain that followed the ash-fall.
8 June B8 6, p. 296 Three columns of fire were seen rising up through the darkness from the central part of Síða, or to the north of the farm
Mörtunga that is inland from the farm Prestbakki. [The trace of a line from Prestbakki to Mörtunga is the direction toward
the westernmost craters on the Laki cone row located near Hnúta Mountain, see Figure 2].
8 June B9 9, p. 1 During that time large smoke and steam columns were constantly seen in the wastelandscalled Síða highlands to the north
of these districts: Síða, Landbrot, Meðalland, Álftaver, NW of Fljótshverfi, but NE of Skaftártunga. There were three
columns of fire of which the westernmost and the northernmost was the largest one. The former one was seen in so called
Úlfarsdalur Valley, just east of Skaftá River gorge according to trustworthy documents; the other two were seen close to
the Hverfisfljót River gorge. These three before mentioned columns of fire, when they were highest, accumulated into one
and reached such height that they were seen from more than 34 miles away [Danish mile = 7532.48 m; distance = 260 km].
Most of the time these fires were screened out of sight by mist and smoke. [This description is a brief summary of the
eruption by S. Hólm. The three columns of fire indicate to the three eruption sites seen from the Fire districts and referred
to as Útnorðursgjá, Landnorðursgjá and Austur-Landnorðursgjá by Steingrímsson].
8 June B10 9, p. 6 On 8 June the before mentioned eruption began ...huge smoke and ash cloud emerged with great force, cracking sounds and
rumbles, carrying an enormous amount of sulfurous material, sand, ash and scoria. Part of the tephra was dispersed over
the pasture and the mountainous wasteland, but another part of it was carried over the rural settlement according to how
the wind was blowing each time. The air was filled with such an amount of sulfurous haze and sandy dust that it was often
impossible to read and write during mid-day. In those districts that where darkened by the sandfall, which partly consisted
of volcanic ash and glowing pumice and partly of ink colored mud-like material and hairs, some of which were rolled up
into small bundles, while others resembled wreaths and thosecaused the greatest damage. On June 8, the above described
ash deposit was dispersed over Síða and some of the other districts and this recurred number of times. On the third day of
the eruption the columns of fire, which continuously rose from the fissure, were seen clearly as numerous flames and at the
same time strong rumbling and cracking sounds were heard from the same place.
8 June B11 10, p. 13 In the morning of Whitsunday, 8 June, the weather was still calm and bright. But around 9 a.m. a large black plume was
seen rising into the air north of here and spread over the Síða district. This plume was not seen from the farms, aligned
along the east-west trending scarp bordering the Síða district, until it was almost over the district. However a few people
from the Landbrot district on their way to church at Kirkjubæjarklaustur saw many columns of smoke rise up over the
mountains in the north and how they accumulated into one to form the dark sandy plume.
9 June B12 2, p. 58 ...same weather [as on 8 June] with southeasterly breeze, the dark sandy plume was now seen clearly in north above the
mountains.
9 June B13 4, p. 9 ...the plume rose continuously. That night heavy rainfall came down from it.
10 June B14 2, p. 58 ...torrential rain from the plume that now continuously rose higherand higher by the day. The rain water was loaded with
before mentioned sandy ash and hairs, light blue in colour and smelled like a mixture of nitrate and sulfur.
40 JÖKULL No. 53, 2003

The 1783–1785 Laki-Grímsvötn eruptions
Table B: Explosive activity, continued.
Date Index Source Quotation
10 June B15 4, p. 9 ...heavy overcast with pungent/acrid rainfall, which caused unbearable smarting in the eyes and on bare skin upon contact
and dizziness in the head. Some raindrops burned holes in the dock leaves as they fell on them and scorched the skin in
newly sheared sheep.
10 June B16 10, p. 14 ...separatecolumns of fire were seen for the first time rising up above the mountains in the north; the ashy cloud rose higher
and higher every day and earthquakes along with rumbling and cracking increased day by day. [Magnús Stephensen did
not witness the Laki eruption and his descriptions of the eruption are based on other eyewitnessaccounts. The description
is very similar to the one Steingrímsson gives for 8 June. It appears that the date 10 June is a mistake by Stephensen.]
11 June B17 4, p. 9 Snowfall and snowdrift in easterly wind, that was derived from the plume.
11 June B18 5, p. 274 ...onthe fourth day of the eruption I [Jón Eiríksson at Ljótarstaðir in Skaftártunga] saw two additional columns of fire north
of the first one and it seemed to be positioned in Skaftárjökull glacier. [The line of sight from Skaftártunga to the eruption
site is such that a northeast shift appears to be toward the north.]
11 June B19 8, p. 420 That night, right after we had put up our tent again, it started snowing and when we had slept through part of the night,
...we were woken up by three enormouscracking sounds so we saddled our horses in a hurry and rode backto the district
in a snow drift and dense ash-fall.
11–14 June B20 7, p. 281 As one could clearly see three separate columns offire and smoke it is possible that the fire emerged from more than one
source, which opened about the same time. These columns of fire and smoke were easily seen from the Rangárvellir district
during 11, 12, 13 and 14 June. After that came a haze containing dust particles, which did not only spread all over the
southern Iceland but also eastern, western and northern Iceland, and blocked the view.
12 June B21 10, p. 15 A huge dark volcanic cloud emerged in the NW and dispersed a great volume of ash, scoria, sulfuric material and greyish
hairy material [over the Síða district]; the land was covered by badly smelling and tormenting smoke which blocked the
sunlight; the sun was seen through this dense smoke and sulfuric mist as a blood-red indistinct globe. Earthquakes and
cracking sounds were frequent and the eruption site was seen through the dense cloud of smoke and mist as numerous
columns of fire north of the mountains bordering the Síða district. [Stephensen contradicts his source here. Steingrímsson
does not mention tephra-fall on 12 June in any of his descriptions. However, Stephensen’s description is similar to that of
Steingrímsson from 14 June both in style and content].
12 June–21
July B22 3, p. 70 This absurdity and enormity continued with thunder and lightning in the air, but earthquakes and din in the earth from 12
June to 21 July. This activity was west of and within my [Steingrímsson] parish. When activity dwindled on this eldgjá
[Útnorðursgjá] another [Landnorðursgjá] began to erupt east and north of us and desolated by ash-fall a whole parish,
named Kálfafell [i.e., Fljótshverfi].
13 June B23 2, p. 59 ...loud claps and cracking sounds were heard from the eldgjá, the column of steam was so high that it was seen from Sel-
vogsheiði in Gullbringusýsla and surrounding districts to the west, [i.e., Reykjanespeninsula, distance from Laki fissures
∼250km.]
13 June B24 4, p. 9 The ash and the steam plume was so high that it was seen all over the country, west to moors in Gullbringusýsla ...the
weather was calm and clear with wind of southwest. Loud claps and cracking sounds were heard here from the northwest,
along with earthquakes, also boiling and noises like commonly heard around waterfalls or as many forge hearths were
blown at the same time. This noise and rumble were continuously heard from this direction for the next three weeks.
13 June B25 2, p. 59 This same day a few men climbed the mountains bordering the Síða district to explore and locate these fires. They saw
20 columns of fire in the so-called Úlfarsdalur Valley, which is situated north of the Síða highlands, and two of them were
indescribably large and they saw 7 other columns in the hillocks north [correct is to the east] of the valley. [Total of 27 fire
fountains were seen in Úlfarsdalur Valley, which is actually the numbers of openings on the two western most fissures of
the Laki cone-row.]
13 June B26 4,p 12 The activity was intense north of the mountains bordering the Síða district, with cracking and crashing sounds, fire and
haze, along with earthquakes such that no one was certain whether the settlement was going to be saved. Therefore
3 farmers, who lived at the farm Mörtunga, climbed up to the highest lookout on Kaldbakur mountain to evaluate the
progress of these fires in the pasture. They saw, as it appeared to them, fires in Úlfarsdalur Valley and they could recognise
22 columns of fire emerging from the fissure.
14 June B27 2, p. 60 ...huge amount of hairy sand was dispersed all over [Síða], but in early evening a heavy rainfall occurred from the plume
so pestiferous and smelly that breathing became difficult for people with respiratory problems.
14 June B28 2, p. 60 The before mentioned hairy sand-fall and sulfurous rain caused such unwholesomeness in the air and the earth that the
grass became yellow and pink and withered down to the roots. The animals that wandered around the fields got yellow
colored and sore feet, and yellow spots were seen on the skin of newly shorn cheeps, which had died.
14 June B29 4, p. 10 ...huge volume of ash, which contained more hairs than previous ash-fall on 9 June [correct date is 8 June], which were
bluish black and glittering, their length and thickness was similar as a hair of a seal (it has been reported that they contained
iron and copper mixture); and they completely covered the ground. Where the hair fell on bare ground it was rolled up
into elongated, hollow bundles by the wind. In the night that same day a heavy rain fell from the plume, despite the
southeasterly breeze, which had muddy or light-blue colour and very bitter and smelly so that people with respiratory
problems could hardly breathe and were dizzy. All summer birds and nesting birds fled, their eggs that were left behind
were hardly edible because of their bitter and sulfurous taste.
15–17 June B30 2, p. 60 ...thunder continued to be heard from the cloud...
16 June B31 4, p. 12 The din and the cracking, the fires and the stench along with the earthquakes to the north of the mountains bordering Síða
was so frenzied, that no one was sure whether this settlement was safe.
JÖKULL No. 53, 2003 41

Thordarson et al.
Table B: Explosive activity, continued.
Date Index Source Quotation
17 June B32 4, p. 13 ...the column of fire rose up so high, from the before mentioned Útnorðurs Úlfarsdalsgjá from where a continuous boiling
sound could be heard, that sparks of fire actually advanced forward between Lambatungur, located at the west end of
Kaldbakur, and Geirlandshraun, which is a elongated hill north of the Síða district, trending east-west and stretches so
far west that it reaches beyond the farm Skál. The fire column was seen [from Prestbakki farm] above the Geirlandsheiði
moor. [This description refers to the rootless eruptions in the Leiðólfsfell area (Thordarson et al., 1998)].
19 June B33 4, p. 14 Southerly winds that directed the main eruption plume northwards towards the glaciers and frequent thunders were heard
from the eruption site.
21 June B34 2, p. 62 ...large amount of ash fell here in the Síða district, followed by sleet and snowfall so the mountains became white. [This
tephra fall occurred in strong easterly wind, suggesting origin at the Grímsvötn volcano].
21 June B35 13, p. 593 The equinox, 21 June, was followed by rainy and foggy weather. The face of the Earth became white. Grass withered
down to the roots like it had been burned. Milking cows and milch ewelost their yield. Due to the haze and the mist the
sun was blood red in colour, especially mornings and evenings. Greyish ash or sand was seen deposited on thin boards
and white paper which was laid outside. To people it appeared to be sulfur, which had been dispersed over the ground,
spoiled and burned the grass and was unwholesome for animals as well as humans. This sulfuric deposit is most likely
derived from the abnormal eruption to the east of us. This deposit was mainly dispersed over Northern Iceland between
the Hrútarfjarðará River and Þingeyjarfljót River [Skjálfandafljót], but not over southern and western Iceland as has been
claimed by some. This could be called the grass burning summer or the sulfuric summer.
21 June B36 4, p. 14 ...Wind from the east with rain [in Síða district]...
22 June B37 4, p. 14 ...same weather continued.
23 June B38 4, p. 17 ...loud whistling was heard from Útnorðursgjá.
23–24 June B39 4, p. 14 ...same weather as before [i.e., easterly wind], which directed the plume towards the glaciers [i.e., Mýrdalsjökull and
Eyjafjallajökull glacier], but the same was heard, seen and felt from the erupting fissure as on 19 June.
25 June B40 4, p. 17 ...same easterly and south-easterly wind, at this time the activity from the same fissure [Útnorðursgjá] was so intensive
that the flame was seen through the cloud and so intensive that clasts and sparks were carried by the wind to Þórsmörk
and Tindfjöll in the west. Because of this tephra-fall the inhabitants of Fljótshlíð and Landeyjar thought eruption had
started in the vicinity, but instead turned out to be pyroclasts from here [i.e., Útnorðursgjá].
27 June B41 2, p. 63 ...huge amount of sandy tephra was dispersed over Síða; during the following night and day a strong westerly wind was
blowing and blew of some of the sand [i.e., tephra] off the ground.
27 June B42 4, p. 17 ...wind was from the west and bringing with it a torrential rainfall. We here in the Síða district saw the edge or the
bottom of a huge black sandy plume [rise from Útnorðursgjá], which twisted and spun up over the mountains forming
bolsters. This cloud dispersed sandy ash and pumice-like clasts over Fljótshverfi (the eastern part of Síða), which [i.e.,
the pumice-like clasts] before and after did not reach this district
28 June B43 2, p. 63 ...the wind continued blowing from the west directing before mentioned plume away from us and over the Fljótshverfi
district where heavy fall of ash and pumice-like clasts occurred. This tephra- fall stunted the grass growth this summer
so it could not be cut and the whole of Fljótshverfi became desolated except Núpsstaðurfarm which was located furthest
east.
30 June B44 4, p. 19 ...thunder and earthquakes along with loud noise from Útnorðursgjá.
30 June B45 4, p. 24 ...the steam plume was directed west towards Mýrdalur. [According to Steingrímsson this steam plume came of the lava
flow in the lowlands in front of the Skaftá River gorge.]
in June B46 5, p. 275 We have not got any exact accounts from the Síða district about what has been destroyed, but we know it has been a very
difficult time for the inhabitants, because in Síða the tephra-fall has done much more damage thanhere [in Skaftártunga]
since from the beginning of the eruption the wind has most of the time been out of the west and thus directed the cloud
towards east.
6–7 July B47 4, p. 20 ...still loud noises were heard from the same fissure. [Útnorðursgjá].
9 July B48 2, p. 64 ...once again a huge tephra-fall was dispersed over here [i.e., Síða district], so grassing animals would not eat the grass
of the fields unless the tephra was raked off, this continued for two nights, then heavy rain-fall occurred which washed
off the tephra.
9 July B49 4, p. 20 ...volcanic ash was dispersed all over the Síða district, so the ground became black.
10 July B50 4, p. 20 ...continuing tephra-fall. ...We heard that no tephra fell in the Meðalland district.
11–12 July B51 4, p. 20 Heavy rain and wind [in the Fire districts] so the volcanic ash was washed down in the ground or was blown off such that
the ground was visible again.
18 July B52 2, p. 65 ...here all around us threats from thunder, lightning, creaking and cracking sounds went on such that a minute was not
missed out for two whole days. Never had so much been going on sincethe gjá [Útnorðursgjá] began to erupt...more had
gone on around the Skaftá River gorge and nearby areas, as was clearly visible on the ground [i.e., the soil cover] as later
observations revealed; it was ripped apart and had been thrown around and had undergone amazing transformation. There
[in the vicinity of the Skaftá River gorge] we found here and there fire-blobs, which had fallen down from the air and
burned the grass around them as they chilled andlithified. Some of these blobs were half buried in the ground and shaped
like a cow-dung. Others were shaped like twisted bundles and had pierced into the ground and broken up on impact.
These fire-blobs appeared to weigh around ten pounds or more [these spatter bombs were producedby rootless eruption
within the lava flow]. When these above described phenomena had been going all night and large part of the day; a sandy
and muddy ash was dispersed over the ground here so it blackened. This tephra-fall came from east-northeast and was
different from previous ones in colour and feel. It came from another eldgjá as later became evident. [East-northeast is
the direction to Grímsvötn volcano. See text for further discussion].
42 JÖKULL No. 53, 2003

The 1783–1785 Laki-Grímsvötn eruptions
Table B: Explosive activity, continued.
Date Index Source Quotation
18 July B53 4, p. 22 ...sandy ash was dispersed all over the Fire districts such that in the Fljótshverfi district all grazing fields were destroyed
as far as to the river Djúpá. [Tephra-fall from Grímsvötn volcano.]
19 July B54 2, p. 65 ...the volcanic plume stayed north ofthe mountains bordering the Síðaand was directedtowards west,occasional thunder
was heard. [Continued tephra fall from Grímsvötn?]
20 July B55 4, p. 22 ...same cloudy weather with thunder, lightning, cracking sounds and earthquakes.
21 July B56 4, p. 23 The plume was directed towards the northwest and after this day no loud rumbling sounds were heard from this Vesturgjá,
despite fires continued to burn and flow from it into September.
22 July–2
August B57 2, p. 66 ...occasional heavy rainfall, and during this time columns of fire and smoke with intermittent thunder and rumbling were
observed in the pasture.
24 July B58 4, p. 23 ...intensive fire coloured glare was seen in the sky...
24 July B59 7, p. 281 Fine-grained pumice-like material, which resembled finely cut tobacco,has been observed to fall in many localities, for
example on the Kjölur route along the Blanda River. [The section of the Kjölur route mentioned here is located in eastern
part of Húnavatnssýsla in North Iceland, 250km north-northwest of the Laki fissures. The timing of this tephra fall is
uncertain but it obviously occurred before 24 July].
25 July B60 4, p. 23 Same fire coloured glare.
28 July B61 4, p. 23 ...rain in westerly wind, with sandy ash and a bad smell.
29 July B62 4, p. 25 ...calm weather, for the firsttime rumble and aboiling sound washeard northeast ofKaldbakur mountain, on astrike with
a high mountain called Blængur. The rumble and cracking was not any less than was heard earlier from Útnorðursgjá,
which had dwindled considerably at this stage. Later that day an awesome plume emerged from the site with ash-fall
that was largely dispersed over the Fljótshverfi district and the eastern part of the Síða district, reducing daylight such
that visibility in our homes was greatly diminished. [The line of sight from Prestbakki to Blængur is the direction to the
first fissure east of Laki mountain, labelled 6 on Figure 2.]
30 July B63 4, p. 26 ...thunder, rumbling and cracking sounds were heard almost continuously from all sides.
in July B64 4, p. 25 Near the farm Skaftárdalur on the eastern side of the Skaftá River gorge lava blobs [spatter bombs], which had fallen out
of the air, could still be seen, some were elongated and twisted together like a cow-dung. Some were still in one piece;
others had broken up on impact. [These spatter bombs were produced by rootless eruption within the lava flow, see also
description on 18 July.]
in July B65 4, p. 25 From there [Skaftárdalur farm] columns of fire were seen at two locations until 14 January 1784.
1–3 Aug. B66 4, p. 26 ...same boiling continued to be heard from this gjá [Landnorðursgjá], along with earthquakes, rumbling, thunder and
lightning...
6 Aug. B67 6, p. 297 Still the entire Fljótshverfi district was completely covered with sandy ash [i.e., from the tephra fall on 29–30 July].
9 Aug. B68 2, p. 66s ...intensive thunder and lightning was noticed around the fissure north of Fljótshverfi.
10–17 Aug. B69 4, p. 27 ...the boiling continued to be heard from Landnorðursgjá...
14 August B70 4, p. 27 ...the plume was directed away [from the rural settlement] by south-westerly wind.
<15 Aug. B71 12, p. 279 Two columns of smoke were seen from the farm Húsafell in Borgarfjarðarsýsla, which rose above the eastern part of
Kaldidalur Valley. [This description is in a letter written by Ólafur Stephensen on 15 August 1783. Here he is describing
the eruption columns of the Laki eruption. From this location the columns must have been viewed over the mountains at
the southern end of Langjökull glacier and to be seen from Húsafell the plume must have been at least 8.5km high and
most likely considerably higher].
17–23 Aug. B72 4, p. 28 ... the same boiling sound was heard from Landnorðursgjá and showers, containing sandy ash, fell frequently [in the
Síða district].
23–31 Aug. B73 4, p. 28 ... activity continued as previously described, but now again intense tremor and cracking was felt from Austurgjá.
in August B74 4, p. 27 When the smokes coalesced and the fumes in them became too heavy, a stinking and sandy rain fell out of the cloud with
lightning and thunder. This occurred despite that clear and bright weather was around us.
in August B75 4, p. 27 When the activity dwindled, which commonly was at its peak during new and/or full moon, the fields recovered to some
degree.
1–7 Sept. B76 4, p. 28 Rain and acrid rain, fog and mist, thunder and lightning occurred frequently that week [in the Fire districts].
7 Sept. B77 7, p. 281 ...in the morning of 7 September fine-grained ash fell on a boat west of Vestmannaeyjar, so the sail and cloth covers
became dark grey in colour in 2–3 hours.
14 Sept. B78 4, p. 29 ...heavy ash-fall in easterly wind, which only reached as far west as Geirlandsá. [Geirlandsá River originates in Lauf-
fellsmýrar up in the Síða highlands flowing almost directly south to the central part of the Síða district. Thus it is
impossible that this ash-fall came from the Laki fissures because the wind was blowing from the east at the time. The
only possible eruption site which could have produced this ash and caused this type of dispersal is Grímsvötn.]
14–26 Sept. B79 4, p. 29 ...at nights the glare from the eruptive fissure reached up to the middle of the sky, especially in two places as it was seen
from Prestbakki. From Vesturgjá the glare was seen between Lambatungur and Kaldbakur mountain, but from Austurgjá
between Kaldbakur and Vothamrar.
26 Sept. B80 4, p. 29 An additional column of fire and steam, which had been seen few times before, appeared far east- northeast from Síða,
due north of Lómagnúpur, but northwest from Öræfi. This activity was ongoing until 24 October. That day intensive
earthquakes and strong sulfuric stench was noted from east-northeast, which indicated that this event was not all over yet
as became evident later... [These directions leave no doubt that Austur-Landnorðursgjá is Grímsvötn volcano].
25 Oct. B81 4, p. 30 ...a high column of fire ascended into the air from there [Grímsvötn] and was followed by an intensive lava surge [from
Útnorðursgjá] along with rumbling and claps for the next five days.
JÖKULL No. 53, 2003 43

Thordarson et al.
Table B: Explosive activity, continued.
Date Index Source Quotation
24 Nov. B82 4, p. 31 The activity was still so intensive in east-northeast,that the column of fire were seen above mountain Kaldbakur. In early
December the activity started to dwindle and the glare, which had been seen almost daily above the three gjá’sor vent
areas, diminished.
in 1783 B83 13, p. 593 The spring 1783 was reasonably good, but with occasional frost. More common was sunshine and clear weather until
Trinitatis [15 June]. Thereafter a mist, haze and fog were seen in the air everywhere, so the sun was hardly seen even in
clear weather. Right after equinox, 21 June, rainfall along with fog came over. Then the face of the earth became white.
The grass withered down to the roots, like it had been burned. Milking cows and sheep lost all milk. The sun appeared
blood red, especially in the mornings and evenings, due to the mist or the haze. Gray sand had deposited on a few thin
boards that had been lying outside and on spread out paper. It appeared to most people that it was sulfur, which was
spread over the ground and damaged the grass and had unwholesome effects on animals and humans. Surely this sulfur
might be from this ongoing and abnormal eruption east of us. The sulfuric deposit was mainly dispersed over the area
between the rivers Hrútarfjarðará and Þingeyjarfljót [Skjálfandafljót], but not over southern and western Iceland as has
been previously said by some. This summer could be called the summer of the Grass-burns or the Sulphuric summer.
...The mist and the haze, and even acrid rainfall, was observed here off and on well into the fall, but the stench people
could smell well into the winter. Brigs and small fishing vessels came to Iceland this summer. Some of them went astray
because of the haze and the smoky mist that spreadall over.
in 1783 B84 14, p.456 Because of this eruption sand and ash were dispersed all over the island. The ash-fall, the haze and the mist was so
intensive that the normal visual field [i.e., view between farms] was reduced considerably and the sun was seen blood
red in color. This was most effective shortly after the mist was first observed or around Trinitatis. It has been verified
that this volcanic dust fell in Bergen [Norway] though without any damage to vegetation or livestock.
January
1784 B85 4, p. 32 Bad smell and odour was often noticed throughout the year [1784], when the wind was blowing from the direction of the
crater area. Fire was still seen Austur-Landnorðursgjá. From Kirkjubæjarklaustur it was seen just north of Þverárfjall
mountain. It was more clearly seen from the Öræfi district, due to its location. [These directions give Grímsvötn as the
eruption site.]
8 April
1784 B86 4, p. 34 This jökulhlaup, observed in the river Súla on 8 April, which was Maundy Thursday, ... the putrid smell and the stench
which followed was so strong, that it could be sensed in every nook and corner inside the farms and never before has
anything like it been reported here. A flash flood was observed in the river Skeiðará at similar time. Jökulhlaup’srecurred
occasionally until the Midsummer Eve. ...The easternmost column now became very black, but since most of this time
the wind was blowing off the sea [southerly winds] the tephra fell mostly on the mountainous desert in the north. Twice,
a small amount of ash fell from the cloud here in Síða.
Summer
1784 B87 10, p. 31 The moor, were many farms in the Síða district had their summer dairy farms due to good pasture lands for the grassing
livestock, was now scorched or completely covered by volcanic sand and ash. As we travelled, further north into the
pastures the thicker the tephra cover became, reaching 4–6 inches in thickness. Some compaction of the deposit had
occurred from the year before and at few places some plants and straws were seen to sticking through the tephra cover.
The once grassy Síða highlands was now completely covered with still thicker tephra sheet and no plants were observed
at all.
January
1785 B88 4, p. 43 The year 1785 started with severe frost, such a piercing and profuse cold caused by the fiery fumes commonly seen as a
bluish mist in the air on clear days.
Spring 1785 B89 4, p. 43 The spring was moderate, but in the spring loud rumbles were heard from the volcanic plume [i.e., Austur-Landnorðurs-
gjá], most intensively on the 4 and 26 May, shortly after a jökulhlaup in the Núpsvötn River. [This was the last time fires
were noticed at Grímsvötn.]
Summer
1794 B90 11, p. 558 Pálsson’s observations at Lauffellsmýrar bogs: Prior to the eruption this area was impassable bogs/mires, but during the
eruption they were filled up by black scoria so now it could be passed on horses without any trouble at all. The scoria
had completely disappeared into the grassroots, so the area has gained back its boggy appearance and with time it will
become the same impassable bogs as it was before.
Summer
1794 B91 11, p.560 Pálsson’s observations on the tephra cover at Galti mountain: As I passed over, I measured the thickness of the scoria
cover which was still preserved in the slopes next to the lava flow, eleven years after the eruption. It was usually around
6 inches thick, even on a slope where one expects that large portion would have been removed by wind and melt water
action during this long interval [that has passed since the eruption came to an end].
Summer
1794 B92 11, p. 567 Pálsson’s observation during his hike across the lava to the Laki cone-row: We were now at the place where I [Sveinn
Pálsson] had decided to walk across the lava flow [to the craters] and the lava was about quarter of a mile wide [1/4
Danish mile = 1884 m]. Finally we came across a deep and wide channel with a smooth bottom, which the lava had
flowed down during the final stage of the eruption and cooled down without breaking up. We managed to cross the
channel and the western branch of the lava flow without any mishaps and reached the before mentioned smoky hillocks.
...After a brief rest, we climbed the closest knoll, which appeared to be the highest one. ...I was very surprised when I
reached the top and saw that the knoll was hollow inside, similar to an Icelandic fold or ruins were the walls are the only
thing left standing. The floor of the knoll was at similar level as the base of its slopes. It was made of solidified lava, but
the walls on the inside looked like they had been built up of incinerated lava blobs and the brink was very thin with lot of
concentric cracks. ...In one of those lava made fences was an opening, at the same level as the floor and made up of same
material, which led to a narrow channel extending towards the large channel mentioned above. This knoll contained two
such craters or bowls, where one had opened a channel towards SW and the other towards NE. The diameter of these
bowls was about 30 fathoms at the top and 10 at the base. ...I investigated some other knolls at this location and they
were all similar to the one already described. On the outside they were mainly built up of reddish gravel-size scoria, on
the inside however they were made up of solidified lava and from each of them a channel extended out into the main lava
field. [The craters described here are located on fissure 3 (Figure 3)].
44 JÖKULL No. 53, 2003

The 1783–1785 Laki-Grímsvötn eruptions
Table C: Lava flows
Date Index Source Quotation
in June C1 3, p. 69 This lava flow [the Laki lava flow] comes from one of the boggiest dells here in the Síða highlands, to the north of us.
in June C2 1, p. 273 The northwestern sector of the highlands, which used to be grassy fields, is now covered by lava. Not only did we loose
a good sheep walk, but also the fields where angelica roots and Iceland-moss were be picked for food and are of vital
importance for the inhabitants.
9 June C3 4, p. 9 The Skaftá River, which ran eastward at the foot of the Síða scarp, ...began to dwindle considerably.
10 June C4 4, p. 9 The channel of the Skaftá River dried up completely, apart from the local tributaries that flowed into it.
10 June C5 10, p. 14 ...To everybody’s surprise the Skaftá River dried up and disappeared on this day.
11 June C6 2, p. 59 ...it was noticed that Skaftá River had dwindled considerably apart from the tributaries, which flowed into it from the
mountains bordering the Síða district.
11 June C7 8,p. 420 ...on the third day after Whitsunday three of us from the farm Skál went woodcutting up in the mountains. Then we
noticed that the great Skaftá River had dried up, therefore we cut the trees into the gorge. North of us further up the
gorge we saw a huge smoke or steam cloud. Late that day we went to explore this further and then we noticed, just a
short distance from we were, that the whole gorge was filled with lava and its sides were like glowing iron. The rocks,
both the glowing and unburned ones collided in the air causingloud cracking sounds.
12 June C8 2, p. 59 ...the lava surge first emerged out from the Skaftá River gorge with incredibly loud cracks, claps, din and rumble. The
discharge of the lava was similar to that the Skaftá River, which is one of the larger rivers in Iceland and was at a high
stand at this time. When the lava poured into the rivers or quicksand, which happened to be in its way, loud claps and
rattles were heard. To begin with the flow advanced at torrential speed along the main river channel, then it poured over
the older lava flows which were in its path and destroyed most of the fields and pastures belonging to the farm Á. [This
was the first of five lava surges to emerge from the Skaftá River gorge].
12 June C9 3, p. 69 ...out from the gorge, named Skaftárgljúfur [Skaftá River gorge], emerged such a large flood of lava, with the colour of
molten copper in a crucible, that it was similar to the water level in the Blanda River during high stand as it is seen at
your place or in the vicinity of the farm Bólstaðarhlíð.
12 June C10 4, p. 9 ...a lava surge emerged out of the Skaftá River gorge with enormous effluent, cracking, din and rumble. When the lava
fell into quicksand, fresh water springs, or river channels loud claps were heard, just like many cannons were fired off at
the same time.
14 June C11 2, p. 60 The following night the flow of lava out the Skaftá River gorge and along the mountain Árfjall was so intensive that the
ground trembled and vibrated from incessant strokes of lightning and thunder, so almost no one got any rest or sleep
that night. At that time the lava burned up and destroyed all land between Skálarstapi and Nesrof [Nes], along with
Brandaland, which belonged to Kirkjubæjarklaustur. This day the flow reached Holtsgarðar where its advance was
halted for awhile. [This is the second lava surge that emerged from the Skaftá River gorge].
14 June C12 4, p. 11 ...aan awful lava surge emerged from the Skaftá River gorge, such that the whole gorge appeared to be filled with lava
to the brim. This surge completely abolished and destroyed the following farms which belonged to the abbey and the
king. Á in Síða and Nes in Skaftártunga, each worth 12 hundred coins of silver. It also completely covered the old lava
flows between the Síða and Skaftártunga districts, which were largely grown up with brushwood and trees, very useful
pasturelands. This area included Brandaland, a good woodcutting area, belonging to Kirkjubæjarklaustur ...Brandaland
was located westsouthwest of Skálarstapi in the nook where the river turns east along the Síða scarp and bordered by
a branch from the Skaftá River. This surge destroyed the so-called Skálargarðar and Holtsgarðar, where it became
stationary for awhile. Another branch of lava advanced south towards Meðalland, where it flowed first over the Botnar
and Steinsmýri lavas. [This description is under 16 June in the original source, which appears to be a mistake. See text
for further discussion].
14 June C13 8, p. 420 The night before Trinitatis [15 June] the flow of lava was so intense that no one had experienced such enormity and
people thought that this would be their last day alive. At this time the lava flow was 20 fathoms away from the graveyard
at the farm Skál and it was as high as highest sea-cliffs (i.e., Vogastapi)[Vogastapi is at the Reykjanes peninsula. The
cliffs that face the ocean are ∼40m high].
15–17 June C14 2, p. 60 ...during these days the lava flowed at the same rate towards south and southwest from the farm Skál, which was still
standing, the flow crept into the older hollow lavas,caused them to swell up so high that the ones which didn’t observe
it would find it hard to believe. These old crags and boulders were thrown up into the air with a cracking noises as many
cannons were fired off, but on impact a loud claps and rumble were heard. [here the Laki lava is advancing over the
934AD Eldgjá lava].
17 June C15 4, p. 13 ...the column of fire rose up so high, ...that sparks of fire actually advanced forward between Lambatungur, located at
the west end of Kaldbakur, and Geirlandshraun, which is an elongated hill north of the Síða district, trending east-west
and stretches so far west that it reaches beyond the farm Skál. The fire column was seen from Prestbakki farm above the
Geirlandsheiði moor. [This descriptions refers to the rootless eruptions in the Leiðólfsfell area (Thordarson et al., 1998)]
18 June C16 2, p. 61 ...an enormous lava surge emerged from Útnorðursgjá, filled the Skaftá River gorge up to its brims and advanced rapidly
out over the fluvial plain so it looked like it was on fire from one mountain side to another. ...the lava flowed so rapidly
out from the gorge, that it resembled the discharge in the river Hvítá during high stand at Skálholtshamar and the flow
velocity was similar to that in a river during the period of spring-thaw... [The third lava surge to emerge from the Skaftá
River gorge; Skálholtshamar is a cliff atthe bank of the river Hvítá near the old episcopal seatSkálholt].
18 June C17 4, p. 13 ...again a lava surge emerged from the Útnorðursgjá, so the Skaftá River gorge, which was both wide and deep, was
filled with lava to the brim.
19 June C18 4, p. 14 The lava that rushed forward this day, filled the channel of river Skaftá and in 5 days it wasted and destroyed the farm
Hólmar, both farms at Fljótar, Hólmasel and Botnar.
JÖKULL No. 53, 2003 45

Thordarson et al.
Table C: Lava flows, continued.
Date Index Source Quotation
19 June C19 2, p. 62 ...the flow split up, one branch advanced eastward along the mountain [Skálarheiði], but another due south along the
channel of the river Melakvísl. ...part of this lava was so high that it was equal the height of Múli next to the farm Skál
or about 60 fathoms high. In forcing its way along the slopes the flow pushed itself under the soil cover and bundled it
up like a cloth. [One fathom is equal to 1.67 m; 60 fathoms = 100 m].
20 June C20 2, p. 62 ...the lava continued to advance as before and flowed eastwards burning low lying grounds in front of the farm Skál
and all the way east to Krókstorfa. ...The branch which advanced down the riverbed of Melakvísl continued flowing due
south. ...This day a few farmers went to explore what the eruption was doing north in the pasture area. Lava had advanced
eastwards towards the middle of the pasture and destroyed good quality grass fields. Little more than two thirds of the
pasture area was lost under lava along with swan-hunting and bird-nesting grounds (which provided the inhabitants with
eggs), fields where Iceland-moss and roots from angelica were picked, which was part of the inhabitants’ diet.
20 June C21 6, p. 296 ...an exploratory party went as far north into the highlands as possible to investigate the status there. Those who knew
the highlands recognised that the fires emerged from three small lava streams in the northern part of the pasture (almost
“dagleið” to the west from the glaciers), located on a flat fluvial plain north of the Hellisá River. The valley [Varmár-
dalur] where the inhabitants of Síða picked roots was east of the lava streams. Activity increased with loud cracking,
ejecta, ashy cloud, yes a storm emerging from the earth. The lava streams grew steadily and covered more ground as
they moved away from the source until they merged into one main stream, which flowed like molten copper, first into
above mentioned valley and when it was filled, the lava threw itself westwards off the mountains above the southwestern
part of the Síða district, into the Skaftá River gorge. [“Dagleið” was a day’s journey on a horse, ∼25 km].
21 June C22 2, p. 62 This Saturday when the lava passed over the Melakvísl rapids, it ran very rapidly across the Steinsmýrarfljót River and
onto the Stekkjatún grass fields north of the farm Hólmasel.
22–23 June C23 2, p. 63 ...this day and the following day the lava advanced eastwards down the river bed of the river Steinsmýrarfljót and stopped
just west of the farm Efri-Steinsmýri.
23–26 June C24 2, p. 63 ...the lava flowed steadily out from the Skaftá River gorge and earthquakes were felt in between, intensive thunder and
lightning occurred all around us.
24 June C25 4, p. 17 ...men from Skaftártunga district went to explore what was happening to the north of the settlement;. they saw that the
lava had emerged from the channel of Hellisá River, here in the pasture.
29 June C26 4, p. 18 Now, for the third time, an enormous lava surge came from above us [i.e., from north],so the area between Skaftártunga
district and Árfjall mountain was once again like one continuous fire. [This is the fourth lava surge from the Skaftár
River gorge].
29 June–12
July C27 2, p. 64 ...the flow of lava did not dwindle a bit, either day or night. However, the flow was pulsating, sometimes more and
sometimes less.
13 July C28 2, p. 64 ...the lava swelled up and flowed east towards Dælur and Fjarðará River, blocked the river in its gorge, then followed its
channel advancing beyond the promontory. Then part of the lava flowed onto the Dalbær moorlands and covered them
with lava, but bulk of the lava rushed into the gorge and then over the high waterfall Stapafoss, filling the large and deep
gorge with lava to the brim, which continued for four days. At that time the gorge could not contain the lava, which
then advanced over the grass fields of the monasterial farm Dalbær and then up to the farmhouse, which it smouldered
to ashes and then buried it by thick lava.
14 July C29 4, p. 21 ...the same day as the lava advanced over the waterfall [i.e., Stapafoss], the fourth and the last lava surge emerged from
Skaftá River gorge, with intensive boiling, cracking and clapping, and enormous din, like everything was going berserk.
[This is the fifth lava surge from the Skaftá River gorge].
18 July C30 4, p. 22 The activity was most intense on 18 July and I [Steingrímsson] couldn’t help thinking that all hell was breaking loose.
18 July C31 2, p. 65 ...here all around us, such threats from thunder, lightning; creaking and cracking sounds went on that not a minute was
missed out for two whole days. Never had so much gone on since the gjá [Útnorðursgjá] started to erupt.
21 July C32 4, p. 23 ...after this day no loud rumbling sounds were heard from this Vesturgjá, despite those fires were seen and lava continued
to flow from it into September.
29 July C33 4, p. 26 ...this day rumble and boiling sound was first heard northeast of mountain Kaldbakur, in the strike of a high mountain
called Blængur. [Onset of activity on fissure 6 to the northeast of LakiMountain].
30 July C34 4, p. 26 ...thunder, rumbles and cracking sounds were heard almost continuously from all sides.
31 July C35 4, p. 26 ...a steam cloud was seen advancing down the Hverfisfljót River gorge. [In the earlier Eldrit Steingrímsson says this to
have happened on 3 August. See also Thordarson this issue].
in July C36 4, p. 25 ...all that month along with August and September the lava continued to flow out of the Skaftá River gorge, but in late
September the flow dwindled and stopped. At this time sheep and other goods were taken over the lava flow west of
Skaftárdalur, because the lava that was still flowing further up in the pasture area kept the rivers to the north of the
Skaftárdalur farm dry.
3 August C37 2, p. 66 ...people noticed that the water in Hverfisfljót River was getting warmer. The temperature increased steadily, until it
finally dried up.
3–4 August C38 4, p. 26 ...along with flowing lava to the north of the mountains, which dried up Hverfisfljót River.
7 August C39 4, p. 26 ...the first lava surge emerged from the Hverfisfljót River gorge. On 8 and 9 August it advanced down the one of channels
of the river Hverfisfljót, the one which was closest to the eastern margin of Síða district towards the south-southwest.
The lava flowed far out onto the sandur plain, beyond the hill Orustuhóll. ...on the eastern side [of the gorge] it advanced
a short distance beyond Dalshöfði. [This was the first lava surge from Landnorðursgjá that emerged from the Hverfisfljót
River gorge. The lava flow front was between Kaldbakur and Dalshöfði mountains on 7 August but on 9 August it had
advanced beyond Þverárfjall mountain].
46 JÖKULL No. 53, 2003

The 1783–1785 Laki-Grímsvötn eruptions
Table C: Lava flows, continued.
Date Index Source Quotation
9 August C40 2, p. 67 ...that day a lava surge emerged from Hverfisfljót River gorge and advanced like running water out onto the sandur plain,
one mile [i.e., Danish mile = 7532 m] beyond the so called Orustuhóll Hillock and from there due south until it stopped.
[The correct distance which the lava advanced beyond Orustuhóll is 4 km].
10–17 Aug. C41 4, p. 27 ...intensive flow of lava out of the gorge, so it was feared that it might advance over the Fljótshverfi district.
17-23 Aug. C42 4, p. 28 ...the flow of lava had dwindled considerably [from Hverfisfljót River gorge].
1 Sept. C43 4, p. 28 ...the second lava surge emerged out of Austurgjá. [As is evident from the quotation above this was the second lava surge
to emerge from the Hverfisfljót River gorge].
1-7 Sept. C44 4, p. 29 ...the lava flow was so vigorous north of the mountains [bordering the Síða district] that it dried up and consumed all
rivers which used to flow out on the sandur plain.
10–11 Sept. C45 4, p. 29 ...yet another lava surge emerged out of the same gjá and advanced between the previously formed lava streams. [This is
the third lava surge to emerge from the Hverfisfljót River gorge].
14–26 Sept. C46 4, p. 29 ...a great break in flow of lava out from both gorges ...because both Skaftá and Hverfisfljót rivers along with all local
tributaries have reappeared and not caused any considerable damage.
26 Sept. C47 4, p. 29 ...was followed by increasing activity and lava production north of the mountains [bordering the Síða district] and again
dried up all the rivers which had reappeared earlier, so people could walk unhindered across the lava for example next
to the farm Skaftárdalur as mentioned above. [This lava surge dried up the Skaftá and Hverfisfljót rivers again. This
increase in the activity is most likely related to the formation of one of the fissures at Fljótsoddi, because it is between to
the glacial outlets of both rivers, which are separated by less than 2 km. This is the fourth lava surge to emerge from the
Hverfisfljót River gorge].
25 October C48 4, p. 30 ...a high column of fire ascended into the air from there [i.e., Grímsvötn], which was followed by a aweful lava surge
[from Landnorðursgjá] along with rumble and claps for the next five days. This was the last lava surge that emerged
from Hverfisfljót gorge and it was also the most fearful and intensive surge, which the Lord God directed between the
earlier ones.
November C49 4, p. 31 All November small lava pulses flowed onto the eastern lava branch, so rumbles were heard.
in 1784 C50 4, p. 32 Neither I or other trustworthy inhabitants observed any lava flowing out of the two gjá’s, which had produced the bulk
part of the lava flow, after 7 February or when 15 weeks had passed of the winter. This eruption lasted therefore for 8
months.
Table D: Sense of directions in the Fire Districts
Date Index Source Quotation
8 June D1 1, p. 272 After a period of intensive earthquakes eldgjá started to erupt, on 8 June, in so called Síða highlands due southwest of
Úlfarsdalur glacier. [This indicates that southwest was sensed from N215◦to 225◦E].
8 June D2 5, p. 274 On Whitsunday I first saw a fire column and smoke northeast from my home [Ljótarstaðir in Skaftártunga], but north of
Síða, which were thought to be located within the glacier where Grímsvötn are situated. [The direction from Ljótarstaðir
to the craters at Hnúta is N24◦E, i.e., northeast = N24◦E].
8 June D3 6, p. 296 Three columns of fire were seen from the central part of Síða or north of the Mörtunga valley as it was seen from the
farm Prestbakki ...[A line from Prestbakki to Mörtunga is the direction towards the southwestern most craters of the Laki
cone-row. Consequently the direction for northwest is actually N320◦E as sensed in the Síða district at the time of the
eruption].
9 June D4 4, p. 9 But the Skaftá River flowed to east along the Síða escarpment. [This is one of many descriptions where Steingrímsson
says that east was parallel to the scarp. Therefore east = N65◦to 75◦E].
12 June D5 4, p. 9 To begin with the lava advanced along the main channel of river [Skaftá], flowed out over the older lavas [Eldgjá lava,
934 AD], which were on either side of the river extending from the gorge east towards Stapafoss. [The Skaftá River ran
eastward along the scarp and Stapafoss was a waterfall in the river next to the hillock Systrastapi. The line of view from
the lava fields towards Systrastapiis parallel to the scarp. East = N65◦to 75◦E].
16 June D6 4, p. 11 It, [i.e., farm Brandaland] was located west-southwest of the Skálarstapi. Hillock... [This gives west-southwest = N245◦
to 255◦E].
17 June D7 4, p. 13 ...the column of fire rose up so high, from the before mentioned Útnorður Úlfarsdalsgjá [the northwest fissure located
in Úlfarsdalur Valley] from where a continuous boiling sound could be heard, that it actually advanced forward between
Lambatungur, located at the west end of Kaldbakur, and Geirlandshraun, which is a elongated hill north of the Síða
district, trending east-west and stretches so far west that it reaches beyond the farm Skál. The fire column was seen from
Prestbakki farm above the Geirlandsheiði moor. [Here the orientation of Geirlandshraun is said to be east-west, but is
actually N70◦E].
19 June D8 4, p. 14 At this time the main lava stream advanced with flying cinders southeast towards Meðalland, mostly it flowed down the
channel of River Melakvísl which before connected Skaftá River and Botnafljót River. [This gives southeast = N150◦E].
24 June D9 4,p. 15 On before mentioned 24 June, the new lava had built up a high pile, I was standing up on a cliff just north of the farm
Efri-Steinsmýri, looking west over the lava flow and could only see the top of Mt. Hafursey. [This gives west = N265◦E].
JÖKULL No. 53, 2003 47

Thordarson et al.
Table D: Sense of directions in the Fire Districts, continued.
Date Index Source Quotation
27 June D10 4, p. 17 ...wind was from the west and carried with it heavy rainfall. We here in the Síða district saw a huge black sandy cloud
rise from Vesturgjá, which twisted and spun up over the mountains forming bolsters. This cloud dispersed sandy ash
containing some pumice-like lapilli scoria over Fljótshverfi and the eastern part of Síða. This was by far the most
voluminous tephra-fall to be dispersed over these districts. [This tephra fall is derived from the tuff cone to west of Laki
mountain. Thus, with reference to the dispersal direction of the tephra fall, west = N280◦to 300◦E].
13 July D11 4, p.21 ...the flow advanced from the west along the channel of river Skaftá into the narrows above Stapafoss waterfall ...from
there the river continued in a deep channel east-northeast along the Klausturfjall mountain. [This gives eeast-northeast
= N25◦to 35◦E].
29 July D12 4, p. 26 ...this day rumble and boiling sound was first heard northeast of Mt. Kaldbakur, in the strike of a high mountain called
Blængur. [This gives northeast = N350 ◦E].
9 August D13 2, p. 67 ...that day a lava surge emerged out from Hverfisfljót gorge and advanced like running water out onto the sandur plain,
one mile [i.e., Danish mile = 7532 m] beyond so called Orustuhóll Hillock and from there due south until it stopped.
[This gives south = N180◦E].
14 Aug. D14 4, p. 28 The lava was rumbling northeast of Eiríksfell. [This gives northeast = N350◦to 360◦E].
26 Sept. D15 4, p. 29 An additional column of fire and steam, which had been seen few times before [i.e., 18 July] far east-northeast from
Síða, due north of Lómagnúpur, but north-northwest from Öræfi district.
January
1784 D16 4, p. 32 Fire was still seen at the Austur-Landnorðursgjá. From Kirkjubæjarklaustur it was seen just north of the Mt. Þverárfjall.
It was more clearly seen from Öræfi district, due to its location. [The line Kirkjubæjarklaustur to Mt. Þverárfjall gives
east-northeast = N20◦to 30◦E].
48 JÖKULL No. 53, 2003