Available via license: CC BY 3.0
Content may be subject to copyright.
1
1
Jens Esmark’s Christiania (Oslo) meteorological observations 2
1816-1838: The first long term continuous temperature record 3
from the Norwegian capital homogenized and analysed 4
5
6
Geir Hestmark1 and Øyvind Nordli2 7
8
1 Centre for Ecological and Evolutionary Synthesis, Department of Biosciences, 9
Box 1066 Blindern, University of Oslo, N-0316 Oslo, Norway 10
2 Norwegian Meteorological Institute (MET Norway), 11
Research and Development Department, Division for Model and Climate Analysis, 12
P.O. Box 43 Blindern, N-0313 Oslo, Norway 13
14
Correspondence to: Geir Hestmark (geir.hestmark@ibv.uio.no) 15
16
Abstract 17
In 2010 we rediscovered the complete set of meteorological observation protocols 18
made by professor Jens Esmark (1762-1839) during his years of residence in the 19
Norwegian capital of Oslo (then Christiania). From 1 January 1816 to 25 January 20
1839 Esmark at his house in Øvre Voldgate in the morning, early afternoon and 21
late evening recorded air temperature with state of the art thermometers. He also 22
noted air pressure, cloud cover, precipitation and wind directions, and 23
experimented with rain gauges and hygrometers. From 1818 to the end of 1838 he 24
twice a month provided weather tables to the official newspaper Den norske 25
Rigstidende, and thus acquired a semi-official status as the first Norwegian state 26
meteorologist. This paper evaluates the quality of Esmark’s observations, presents 27
new metadata, new homogenization and analysis. The air temperature in Oslo 28
during this period is shown to exhibit a slow rise from 1816 towards 1825, 29
followed by a slighter fall again towards 1838. 30
31
32
33
Clim. Past Discuss., doi:10.5194/cp-2016-60, 2016
Manuscript under review for journal Clim. Past
Published: 20 June 2016
c
Author(s) 2016. CC-BY 3.0 License.
2
1 Introduction 34
The current concern with climate change has increased the interest in early 35
meteorological observation series and evaluation of their quality ( e.g. Bergström 36
& Moberg, 2002; Auer et al., 2007). In a recent paper we analysed the temperature 37
record for the Norwegian capital made 1837-2012 by the astronomical 38
Observatory at the University of Oslo and the Norwegian Meteorological Institute 39
(MET Norway) (Nordli et al., 2015). Previous to 1837 long term observations of 40
the Oslo weather were known to have been made by Jens Esmark (1762-1839), 41
professor of mining sciences at the University of Oslo (then Christiania). A first 42
reanalysis of Esmark’s observations was made by meteorologist B. J. Birkeland 43
(Birkeland, 1925). Our rediscovery in 2010 of Esmark’s original meteorological 44
observation protocols has provided an opportunity to digitize, homogenize and 45
analyze his data with modern methods. 46
Esmark is today mostly remembered for his pioneer ascents of many of 47
Norway’s highest peaks (Esmark 1802, 1812; Hestmark 2009), his discovery of 48
Ice Ages, and his astronomical explanation of such dramatic climate change as 49
caused by variations in the eccentricity of the orbit of the Earth is now recognized 50
as a precursor of the theories of James Croll and Milutin Milankovich (Esmark, 51
1824, 1826; Andersen, 1992; Worsley, 2006; Rudwick, 2008; Berger, 2012; 52
Krüger, 2013). In his own lifetime he was primarily known as a skillful 53
mineralogist and geologist. Throughout his life Esmark maintained a passion for 54
meteorological observation with instruments he crafted himself in accordance with 55
the highest contemporary standards. His main inspiration for this activity were his 56
teachers at Copenhagen University, which he attended 1784-89; first among them 57
the Astronomer Royal, professor Thomas Bugge (1740-1815), who in his 58
observatory tower Rundetårn in the middle of Copenhagen made daily 59
measurements of the weather (Willaume-Jantzen 1896). Esmark also befriended 60
Bugge’s instrument maker, the Swede Johan(nes) Ahl (1729-1795) (Esmark, 1825; 61
Anonymous 1839). In addition Esmark followed the lectures of Christian Gottlieb 62
Kratzenstein (1723-1795), professor of medicine and experimental physics, a 63
‘hands on’ practical man who enjoyed crafting instruments and all sorts of 64
machines (Snorrason, 1974, Splinter, 2007). From 1789 to 1791 Esmark studied 65
mining sciences at the Norwegian silver town of Kongsberg, and after further 66
studies in Freiberg, Saxony and Schemnitz, Austria-Hungary, he in 1798 moved 67
Clim. Past Discuss., doi:10.5194/cp-2016-60, 2016
Manuscript under review for journal Clim. Past
Published: 20 June 2016
c
Author(s) 2016. CC-BY 3.0 License.
3
back to Kongsberg to take up a position as Assessor in the central mining 68
administration (Overbergamtet) of the dual kingdom Denmark-Norway. At 69
Kongsberg he also lectured in mineralogy, geology and experimental physics at 70
the Royal Norwegian Mining Seminar, acting as its temporary Inspector 71
(Headmaster) from 1799, and permanent Inspector 1802-1815. From 1 January 72
1799 he three times a day recorded observations of the Kongsberg weather - air 73
pressure on mercury barometers (in inches and lines), and air temperature in 74
degrees of Reaumur; documented in a series of small notebooks running 75
continuously with some lacunae until 16 September 1810, and rediscovered by the 76
authors in 2010 (Esmark 1799-1810). When Esmark in 1815 moved to the 77
Norwegian capital Christiania (now Oslo) to become the first professor in the 78
mining sciences at the University he continued this habit. At least from January 79
1816 up to and until the day before his death on 26 January 1839 he recorded air 80
temperature and barometric pressure three times a day. The complete set of his 23 81
Christiania observation protocols, long believed lost, was rediscovered in 2010 by 82
the authors, and is now safely deposited in the Norwegian National archives 83
(Riksarkivet) (Esmark 1816-1838). They provide a unique and detailed picture of 84
the weather in Oslo in the early 19th century. From January 1818 to December 85
1838 tables of Esmark’s observations were published every fortnight in the official 86
newspaper Den norske Rigstidende (cf. Appendix A), and he thus acquired a semi-87
official position as Norway’s first state meteorologist. Based on a number of 88
previously unpublished documents (cited as Document 1 etc, with archival 89
location in Reference list) we here present new metadata for Esmark’s 90
meteorological observations from Christiania, and homogenize, reanalyse and 91
evaluate his original data with modern statistical tools to characterize the weather 92
in the Norwegian capital in this period. 93
94
2 Metadata 95
2.1 The location - No. 308, Vestre Rode - Øvre Vollgate 7. 96
Esmark’s observations were made at his home (cf. Esmark 1823b: De ere tagne i 97
min Bopel), and there is no evidence indicating that he changed the location. On 19 98
August 1815 Esmark was registred as owner of property No. 308 in Vestre Rode 99
(i.e. Western Quarter), one of the four old quarters of Christiania town (Document 100
Clim. Past Discuss., doi:10.5194/cp-2016-60, 2016
Manuscript under review for journal Clim. Past
Published: 20 June 2016
c
Author(s) 2016. CC-BY 3.0 License.
4
1). It was a modest one-and-half storey house built late in the 18th century with an 101
adjoining a garden. Esmark’s continued residence at this address until his death is 102
documented in annual censuses and tax protocols (Document 2 & 3). Property No. 103
308 was situated on the north-western side of the street Øvre Vollgate (Øvre 104
Woldgaden), laid out literally on what used to be the outermost western rampart 105
(voll) of nearby Akershus Castle and Fortress (Fig. 1). It was a natural rock 106
promontory above a meadow to the west where the poor fishing village Pipervigen 107
would develop later in the 19th century, today the site of Oslo Town Hall. In 1815 108
Øvre Vollgate constituted the south-western limit of Christiania, a town with only 109
about 15000 citizens (Myhre 1990). Until 1814 the main administration centre of 110
the dual kingdom was in Copenhagen, but with Christiania in that year acquiring 111
the new parliament and government after the separation of Norway from Denmark, 112
the town expanded rapidly. When street numbers were introduced, Esmark’s 113
property was numbered Øvre Vollgt No. 7. The present Øvre Vollgate 7 – an 114
office highrise – comprises previous numbers Øvre Vollgate 3, 5 and 7. 115
Esmark’s property No. 308 and all neighbouring properties were measured 116
and mapped for the new matriculation of Christiania in the summer of 1830, and 117
thus we have very precise data on his house and the surrounding properties at the 118
relevant time (Document 4). The whole property roughly constituted an elongated 119
rectangle, approximately 14 m x 60 m (Fig. 2). The unit used in these 120
measurements was the ‘Norwegian alen’ (Norsk alen), determined by law in 1824 121
to be 62.75 cm. It was divided into two feet, each divided into 12 inches, each 122
divided into 12 lines. No. 308 was measured to 2026 square alen, of which the 123
house (including a yard) was 733 ½ and the garden 1292 ½ square alen (1 square 124
alen = 0.3937 m2). Thus the whole property was ca. 800 m2, and the house 125
(including yard) ca. 290 m2. The house had a 22 alen 6 inch (ca. 14 m) long 126
façade towards the street Øvre Voldgate, constituting the south eastern border of 127
the property, with windows, doors, and a gate leading in to the back yard (Fig. 3). 128
Øvre Vollgate street runs from SW to NE at an angle of roughly 32o NE (400 129
degrees). At the back the house surrounded a small yard, with a narrow passage 130
opening out to the garden in the NW. As it would have been hazardous to place the 131
meteorological instruments on the street-side of the house, where passers-by could 132
have tinkered with them, it is almost certain that they were placed in Esmark’s 133
back yard, a well guarded space. When the house was finally demolished in 1938, 134
Clim. Past Discuss., doi:10.5194/cp-2016-60, 2016
Manuscript under review for journal Clim. Past
Published: 20 June 2016
c
Author(s) 2016. CC-BY 3.0 License.
5
it was in such bad condition that the Oslo city health authorities demanded the 135
whole property be sprayed with hydrocyanic acid and that none of the fungus-136
infected material be used for construction elsewhere (Document 5). 137
Esmark’s garden on the NW side of the house and court yard was a 138
continuous slope, dropping ten alen (6,25 m) down along 66 alen length towards 139
Pipervika. Here it was most probably limited by a fence towards the Præste Gade 140
street which later changed name to todays Rosenkrantz gate. In 1841, a couple of 141
years after Esmark’s death, most of this garden was indeed sectioned out and sold 142
to form the new property Rosenkrantz gate 26. In Esmark’s time, however, the 143
promontory remained an open garden space. His neighbours on both sides (No. 144
307 and No. 309) had the same arrangement of house and garden, with facades to 145
Øvre Vollgate and gardens sloping down on the back to Præstegaden (Document 146
6). To the north of the lowermost part of Esmark’s property was an open space 147
called Jomfru Wold’s Løkke (No. 368). South of this lower part of the garden was 148
the street Pipervigbakken, leading down from Rådhusgaten street passing by the 149
outer ramparts of Akershus fortress and Castle. The sea with Pipervika bay 150
(Piperviks Bugten) was less than 200 m south of Esmark’s garden. His garden was 151
not an entirely constant environment. In 1823 for instance, he received several 152
fruit trees from a Danish friend which he planted in the garden (Document 7). 153
It was a modest residence for a professor, situated in a comparatively poor 154
part of town, with mainly craftsmen, tradesmen and artisans in the neighbourhood 155
(Myhre 1990: 40). Here Esmark, a widower since 1811, moved in with his three 156
sons Hans Morten, Petter and Lauritz, a maid and a manservant (Document 2 & 3). 157
His daughter Elise resided with her grandparents in Copenhagen, but later returned 158
to Norway to take up residence in No. 308. 159
160
2.2 The observers 161
The great majority of the Christiania observations were made and noted down by 162
Esmark himself who has an easily recognizable handwriting. His position as 163
professor in the mining sciences did however sometimes cause him to leave town 164
on short or long field excursions, some lasting several months. He was away from 165
Christiania on long voyages in 1818 (Hallingdal), 1819 (Kristiansand); 1822 166
(Bergen), 1823 (round-trip south Norway), 1826 (Setesdalen), 1827 (Trondhjem) 167
and 1829 (Copenhagen). In his absense his sons seem to have been instructed to 168
Clim. Past Discuss., doi:10.5194/cp-2016-60, 2016
Manuscript under review for journal Clim. Past
Published: 20 June 2016
c
Author(s) 2016. CC-BY 3.0 License.
6
continue daily observations, and there are extremely few missing data points. The 169
oldest son Hans Morten Thrane Esmark (b. 1801) in 1825 became a chaplain in 170
Brevig and moved from Christiania; Axel Petter (b. 1804) became a sailor and was 171
often away from home; Lauritz Martin (b. 1806), later a professor of zoology at 172
the Christiania university, and daughter Elise Cathrine (b. 1800) remained at home 173
until Esmark’s death. The sons evidently did not fully share their father’s passion, 174
and although instrument readings were meticulously maintained, the qualitative 175
notes on weather are often restricted to a single word in Esmark’s absence. A 176
claim (Birkeland 1925: 5) that the botanist Martin Flor performed the observations 177
in Esmark’s absence has not been substantiated, and anyway Flor committed 178
suicide in 1820. 179
180
2.3 The hours of day 181
Esmark’s Christiania observation protocols do not indicate the precise hours when 182
the observations were made. The columns are given as morning, noon and evening 183
(Morgen, Middag, Aften). A note on the first published table in Den norske 184
Rigstidende on 24 January 1818, also says Morgen, Middag og Aften without 185
further specification (Fig. 6). In a summary table of 15 years (1818-1832) 186
published 1833 Esmark is more explicit: “The barometer observations have been 187
made daily in the morning, afternoon and evening; in later years at 8 ½ o’clock 188
morning, at 3 ½ o’clock afternoon and 9 ½ o’clock evening; thermometer 189
observations at the same times in the afternoon and evening and in the morning 190
with the help of the night thermometer. From this the middle hight is taken.” 191
(Barometerobservationerne ere dagligen gjorte om Morgenen, Eftermiddagen og 192
Aftenen; i de senere Aar Kl. 8 ½ Morgen, Kl. 3 ½ Eftermiddag og Kl. 9 ½ Aften; 193
Thermometerobservationerne paa samme Tider om Eftermiddagen og Aftenen og 194
om Morgenen ved Hjælp af Natthermometret. Heraf er taget Middelhøiden.) 195
(Esmark 1833: 235). Thus 8.30 AM, 15.30 (PM), 21.30 (PM). The hour 3 ½ PM 196
probably coincided with Esmark’s return to his house from the lectures at the 197
University just a few blocks away. The phrasing “in later years” suggests that the 198
hours had not been constant throughout the whole series. This problem we analyse 199
further below. Also that a night-thermometer (for measuring minima) was 200
introduced some time after the start of the series. 201
202
Clim. Past Discuss., doi:10.5194/cp-2016-60, 2016
Manuscript under review for journal Clim. Past
Published: 20 June 2016
c
Author(s) 2016. CC-BY 3.0 License.
7
2.4 The instruments and their position 203
In a note to his first table presented in the journal Den norske Rigstidende, on 24 204
January 1818, Esmark provides a few details of his measurements: “The 205
observations are made 34 Rhinelandic feet [i.e. 10.68 m] above the sea, and are the 206
middle value of observations made morning, noon and evening. The barometer 207
heights are corrected as they would have been if the barometer was subject to a 208
temperature of 0o. The thermometer hangs freely against north.” (Observationerne 209
ere anstillede 34 Rhinlandske Fod over Havet, og ere Middeltallet af 210
Observationer, anstillede Morgen, Middag og Aften. Barometerhøiderne ere 211
corrigerede saaledes, som de skulle være, dersom Barometret havde været udsat 212
for 0o Temperatur. Thermometret hænger frit imod Nord.) (Fig. 6). Esmark also 213
notes for these (average?) data that “The barometer height is reduced to 0o R. If 214
one wants it reduced to sea level, one must add a line or 1/12 of an inch to its 215
height, so that the barometer height at sea level becomes 28.1,20 in French 216
measure.” (Barometerhøiden er reduceret til 0o R. Vil Man have den reduceret til 217
Havets Overflade, maa Man til den anførte Høide lægge en Linie eller 1/12 Deel 218
af en Tomme, saa at Barometerhøiden ved Havets Overflade bliver 28.1,20 i 219
Fransk Maal.) (Esmark 1833: 235). 220
221
Thermometers. Esmark all his life used the Reaumur scale; “R”. The precision of 222
his Reaumur thermometer was 1/2 of a degree. On a table of averages for the years 223
1816-1822 Esmark notes: “The thermometer observations are made in shadow in 224
free air with a Reaumur thermometer, which boiling point is determined at 28 225
inches 2 lines (French measure) barometric height.” 226
(“Thermometerobservationerne ere gjorte i Skyggen i fri Luft med et Reaumurs 227
Thermometer, hvis Kogepunkt er bestemt ved 28 Tommers 2 Liniers (fransk Maal) 228
Barometerhöide.”) (Esmark 1823). 229
230
Barometer. Of the barometer used Esmark (1833: 235) states: “The barometer is a 231
simple barometer, the tube of which is 2 ½ line in diameter and which capsul is 40 232
lines in diameter, and calibrated after a hevertbarometer.” (Barometret er et 233
simpelt Barometer, hvis Rør er 2 ½ Linie i Diameter og hvis Capsel er 40 Linier i 234
Diameter, samt justeret efter et Hævertbarometer.) 235
236
Clim. Past Discuss., doi:10.5194/cp-2016-60, 2016
Manuscript under review for journal Clim. Past
Published: 20 June 2016
c
Author(s) 2016. CC-BY 3.0 License.
8
2.5 The protocols and data recorded 237
Esmark’s Christiania protocols are handmade, folded sheets of white paper cut up 238
and sewn in with a thin grey cardboard cover, one protocol for each year (Fig. 4), 239
23 protocols in all (Esmark 1816-1838). Esmark interfoliated the official printed 240
Almanach for Christiania. This had for each month 16 days on each page, and thus 241
Esmark wrote down his data for 15 or 16 days on the first page of a month and the 242
remaining days from 17 to 28, 29, 30 or 31 on the next page (Fig. 5). The protocols 243
start on 1 January 1816 and end 31 December 1838, only 26 days before his death; 244
altogether 8401 days of continuous measurements. There are only a few small 245
lacunae. Photographs of all the protocols are available at MET Norway (Klimadata 246
samba server, HistKlim skanna dokument), and digitized values might be 247
downloaded from MET Norway’s home page: http://www.met.no. Esmark 248
continued observations in January 1839 until the day before his death 26 January, 249
but these observations are only known through the newspaper Morgenbladet, 250
which had published Esmark’s daily measurements since 1834. 251
Three times a day Esmark recorded temperature to a half degree, and air 252
pressure with one or two decimals (Fig. 5). In the right hand margin he noted the 253
weather (Veirliget) with qualitative terms; see also Esmark (1833). He used a fairly 254
limited number of categories: Precipitation: lidt Regn (a little rain); Fiin Regn 255
(drissle); Regn (rain); Regn Bygger/Bÿgger (showers); Regn af og til (Rain now 256
and then); megen Regn (much rain); Sne (snow); Sne Flokker (snow); Sne Bygger 257
(snow showers). Cloud cover: Klart (clear), enkelte Skyer (a few clouds); tynde 258
Skyer (thin clouds); skyet (cludy); skyer i Horizonten (clouds in the horizon); disig 259
(haze); Taage (fog). The most common category was tykt (thick) which means a 260
grey day with haze, often with precipitation. Wind: Wind direction was usually 261
recorded only once a day, at midday, with categories N, S, V and O, and 262
combinations , e.g. N. O. (nord ost/north easterly). Other: Torden (thunder); 263
Nordlys (northern lights); Flekker i Solen (sunspots); one or two circles around the 264
sun; Høyt vand (high sea level). In June 1818 Esmark introduced a new parameter: 265
precipitation, measured with a rain gauge, and in the June summary, he could 266
announce: “In this month there has, according to the rain gauge, fallen rain to a 267
height, which, if it had been standing, had constituted a height of 1 inch and 9 and 268
7/12 line. The rain gauge is situated 15 feet above sea level.” The low altitude of 269
the rain gauge suggests that it was placed at the lower part of the slope in his 270
Clim. Past Discuss., doi:10.5194/cp-2016-60, 2016
Manuscript under review for journal Clim. Past
Published: 20 June 2016
c
Author(s) 2016. CC-BY 3.0 License.
9
garden. In October 1820 he presented the readers of Rigstidende to his new design 271
for a hygrometer – an instrument to measure air humidity (Esmark, 1820). It was 272
modified from a model developed by John Livingstone, and M.D. from Canton, 273
China, published in the Edinburgh Philosophical Journal in 1819 (Livingstone 274
1819). The general idea was to put a moisture absorbing/releasing chemical 275
substance (Livingstone used pure sulphuric acid, which was also used to produce 276
ice) on one side of a balance, balanced against a weight on the other side. The 277
balance was placed under a glass jar open in the bottom to let air freely flow in and 278
out, and to protect it from precipitation. Esmark made two new hygrometers 279
according to this model. ”Anyone who desires to see these hygrometers, can see 280
them at my house” (“Enhver, som har Lyst dertil, kan see disse Hygrometere hos 281
mig.”)(Esmark, 1820) He had tested them for several months, and thought they 282
could be used by farmers to predict weather change as substitute for barometers. 283
He did not, however, use the hygrometer data for his meteorological tables. For the 284
year 1821 he presented more regular monthly data on precipitation in inches – 285
from 1 May through October – apparently the months without frost. 286
287
2.6 The published tables 288
Starting on Saturday 24 January 1818, with a table presenting weather data for the 289
first half of the month, the semi-official daily Den norske Rigstidende published 290
Esmark’s meteorological observations, which thus acquired an official air. (Fig. 6). 291
It became a regular series, published twice a month – one table for the first half of 292
the month, one for the second half – a total of 24 tables each year, all with the 293
same title ”Meteorologiske Iagttagelser i Christiania [year], anstillede af Prof. 294
Esmark.” (Meteorological observations in Christiania [year], made by Prof. 295
Esmark) etc.. This series running from 1 January 1818 to 15 December 1838 is 296
absent from all previously published bibliographies of Esmark’s works, but in fact 297
runs to no less than 503 published tables (!) (Appendix A). They present 7665 days 298
of continuous observations. In addition comes the two full years of 1816 and 1817, 299
only published summarily by Esmark (1823) but with complete record preserved 300
in the original protocols. The whole year 1818 was summed up on 8 January 1819 301
with means etc., and here Esmark also compared the Christiania data to those 302
obtained by Wargentin in Stockholm, by Bugge in Copenhagen, and (no 303
observator given) in St. Petersburg, Russia. It was not a weather forecast but 304
Clim. Past Discuss., doi:10.5194/cp-2016-60, 2016
Manuscript under review for journal Clim. Past
Published: 20 June 2016
c
Author(s) 2016. CC-BY 3.0 License.
10
rather a weather ‘backlog’, and this may have dimmed their public interest 305
somewhat. The data given in these published tables differ from the raw data of the 306
protocols by being daily averages. For each day he gave the barometric pressure 307
and temperature, averaged from observations made in the morning, at noon, and in 308
the evening (at first without further precision of hour). To calculate these averages 309
he apparently used the formula: 310
𝑇
!=1
4(𝑇
!+2𝑇
!! +𝑇
!!!) (1) 311
where Tm is Esmark’s daily “mean” temperature, and TI, TII, and TIII are the 312
observed temperature morning, noon and evening, respectively. To the tables for 313
the second half of each month, he also appended a note with the mean barometric 314
pressure and temperature for the entire month, and indicated which days had the 315
maximum and minimum air pressure and temperature. The mean temperature was 316
given to 1/100th degree (a spurious precision). The series continued in 1820, now 317
also with the daily wind direction. Esmark evidently trusted only himself to 318
calculate the averages and set up the tables, and thus the readers of Rigstidende 319
sometimes had to wait for months to read the weather for the last fortnight. From 320
1834 Esmark’s observations were also published in the Christiania newspaper 321
Morgenbladet every day, with two days delay, i.e. observations for the 1st day of 322
the month were published on the 3rd etc. This was initiated after Christiania 323
doctors suspected a connection between the weather and the cholera epidemics 324
which struck Norway from 1833 and forward. 325
326
3 Methods 327
328
3.1 Homogeneity testing 329
A homogenous climatic time series shows variations in climate without being disturbed by 330
other factors involved, like changes in the environment, observational procedures or 331
instrument calibration. For the study of climate variations the use of homogenous series is of 332
paramount importance, otherwise the climate analysis might be wrong (e.g. Auer et al., 2007; 333
Moberg and Alexandersson, 1997; Tuomenvirta, 2001). For testing the homogeneity of 334
Esmark’s temperature series we selected the Standard Normal Homogeneity Test (SNHT) that 335
has been widely used for testing of both precipitation series and temperature series 336
(Alexandersson, 1986; Alexandersson and Moberg, 1997; Ducré-Robitaille et al., 2003). The 337
Clim. Past Discuss., doi:10.5194/cp-2016-60, 2016
Manuscript under review for journal Clim. Past
Published: 20 June 2016
c
Author(s) 2016. CC-BY 3.0 License.
11
first version of the test (Alexandersson, 1986) had one step change as the only possibility, 338
whereas in the version of 1997 both double shifts and a trend were possible outcomes of the 339
test. In any year the significance of a potential break is examined. The testing followed the 340
principle of comparing a candidate series (the series under testing) against a reference series. 341
The reference might be series from one or more neighbouring stations. A candidate series 342
might also be observations at one particular time of the day, which are compared with other 343
observation times for the same station. In the latter case we call it “internal testing”. Without 344
contemporary neighbouring stations internal testing is the only possibility. If no significant 345
break occurs the series is considered homogenous. Esmark’s station at Øvre Vollgate 7 as 346
well as other observation stations used in this article are given in Table 1, with their national 347
station number (identifier) and name. Before the analysis started all observations were 348
calculated from degree of Reaumur to degree of Celsius by multiplying Esmark’s Reaumur 349
readings by the factor 1.25. 350
351
4 Results 352
353
4.1 Homogeneity testing 354
For much of Esmark’s period of observation there was no other nearby station in operation so 355
internal testing was the only possibility. The testing was performed both for seasonal (see 356
Table 2) and monthly (see Table 3) resolutions where observations taken in the morning (I), 357
noon (II) and evening (III) were compared with each other. By comparing several test results 358
it was possible to decide at which observation time a shift (inhomogeneity) occurred. Most 359
striking are the huge shifts detected in spring, summer and autumn when the morning 360
observation was involved. The most probable year for the shift was 1827; in particular this 361
was true for the single shift test. Here we apply the common convention to define the shift 362
year as the last year before the shift. We have to conclude that the morning observation is 363
inhomogeneous. A further investigation of the daily observations (not shown) suggested that 364
the change took place within the month of March 1828. 365
When evening observation was tested against the midday observation a shift seemed to 366
occur in 1820 or 1821, most probably in 1821. But this break in homogeneity was much less 367
than that of the morning observation. The shift seems to be absent or very weak during winter 368
so exact dating was impossible. For convenience the end of 1821 was adopted as the year of 369
the inhomogeneity. 370
Clim. Past Discuss., doi:10.5194/cp-2016-60, 2016
Manuscript under review for journal Clim. Past
Published: 20 June 2016
c
Author(s) 2016. CC-BY 3.0 License.
12
Tests including the midday observation revealed no additional shifts than those 371
already detected. The occurrence of the shifts in the tests I vs II and III vs II seemed to reflect 372
shifts either in the morning or in the evening observations. For the winter season a shift in the 373
last part of the series was detected, possible shift years were 1832, 1833 or 1834. 374
The large shift in the morning observation could have masked possible smaller shifts in the 375
series on both sides of this shift. Therefore the single shift SNHT was applied on two different 376
parts of Esmark’s series: 1816.01-1828.02 and 1828.03-1838.12, parts 2 and 3 in Table 2. 377
However, no further shifts in the series were detected. The shifts detected in part 1 in the 378
evening observations of 1821 and in the morning observation in the 1830s for the winter 379
season were confirmed. 380
The reliability of the results was further tested on monthly resolution and also 381
evaluated by comparison with the metadata. Esmark (1833) tells that he uses “a night 382
thermometer” for the morning observation. Our hypothesis is that in Esmark’s terminology 383
“night thermometer” means “minimum thermometer”, and that the introduction of the 384
minimum thermometer is the reason for the shift in March 1828. This hypothesis was tested 385
by studying the difference between Esmark’s evening observation and the morning 386
observation the following day for the three homogenous intervals (see Table 4) (the winter 387
inhomogeneity in the 1830s was ignored). For comparison this was also done for the 388
observations at the modern station Oslo – Blindern. In the earliest interval (row 1) the 389
differences in Esmark’s observations were very much smaller than those from Blindern, so it 390
is impossible that Esmark could have noted the nightly minimum temperature in the column 391
for the morning observation. In the next interval (row 2) the differences are somewhat larger, 392
but far too small compared to Blindern so the same conclusion has to be drawn: no minimum 393
thermometer was in use. However, in the third interval (row 3) the differences are nearly the 394
same as those for Blindern. Even the monthly variations throughout the year are realistic. We 395
conclude that Esmark for the morning observation used a minimum thermometer in the period 396
1828.03-1838.12. Before that he observed temperature in the morning with an ordinary 397
thermometer. If the minimum thermometer was set at the evening observation the notes in the 398
column for morning observation should always be equal or lower than the evening 399
temperature the previous day. In December this is not true for 26% of the observations and in 400
June for 6%. These figures reduce to 6% and 2% in December and June respectively for 401
violations no more than 1ᵒC. In practice different exposure of the two thermometers may 402
violate this logical test, and one should also take into account the possibility of instrumental 403
Clim. Past Discuss., doi:10.5194/cp-2016-60, 2016
Manuscript under review for journal Clim. Past
Published: 20 June 2016
c
Author(s) 2016. CC-BY 3.0 License.
13
errors in Esmark’s thermometers. We may conclude that the percentage of violation is not 404
large enough to contradict our conclusion that a night minimum thermometer was in use. 405
406
4.2 The shift in 1821 407
An inhomogeneity in the evening observation was detected by the homogeneity testing. It was 408
adjusted for by the mean difference between the midday observation and the evening 409
observation on each side of the shift, cf. Methods. The adjustments terms are presented in 410
Table 5. The adjustments are largest in the months where the daily temperature wave in 411
largest, so it is much likely that one reason for the shift was an earlier evening observation 412
time before 1822. If so it seems that the observation was taken at least one hour earlier before 413
this shift. Strictly speaking we know Esmark’s observation times only in 1833, so this result is 414
not in contradiction to metadata. Other factors than the observation times might as well have 415
been involved, as the adjustments in winter is too large to be due to observation time only. 416
417
4.3 A shift in the 1830s 418
significant in winter, was detected by the SNHT double shift as well as the single shift when 419
the time window for the test was 1828.03-1838.12. The shift has the character of a continuous 420
inhomogeneity (Fig. 7). The difference between the evening observation and the morning 421
observation increased quite steadily from 1831 to 1838, whereas it was constant during the 422
years 1829-1831. The explanation may be a change in the observation times. According to 423
Esmark (1833) his observation times were, see Metadata. 424
• Morning: 08:30 ChT = 08:43 CET = 7:43 UTC 425
• Midday: 15:30 ChT = 15:43 CET = 14:43 UTC 426
• Evening: 21:30 ChT = 21:43 CET = 20:43 UTC 427
ChT = Christiania time i.e. local time for Christiania (Oslo), CET = Central European 428
Time, UTC = Universal Time Coordinated. 429
These observation times were for the barometric pressure, but at midday and evening the 430
thermometer were read at the same time as the barometer, but Esmark does not explicitly say 431
that the morning thermometer was read at the same time as the barometer. He also use the 432
term “in the latest” years so we do not know from which year these observation times were 433
introduced or if he continued to use them also in the following years 1834-1838. 434
Our hypothesis is that Esmark has had another observation time for the temperature 435
observations in the morning than for the pressure observations. Pressure was observed inside 436
Clim. Past Discuss., doi:10.5194/cp-2016-60, 2016
Manuscript under review for journal Clim. Past
Published: 20 June 2016
c
Author(s) 2016. CC-BY 3.0 License.
14
the house, but for the temperature observations he had to leave the house for his garden. 437
Esmark might originally have observed temperature and pressure at the same time also in the 438
morning, but with the introduction of the minimum thermometer he could have thought that 439
the observation time for the morning temperature was not important. In spring, summer and 440
autumn he obviously was right in his thinking as minimum temperature occurs earlier than at 441
the morning observation time (8:30 ChT), but in winter the minimum temperature occurs 442
often later in the day as the systematic daily temperature wave is weak. This can explain the 443
changing difference during winter and the stable differences during the other seasons. As 444
Esmark grew older he might have gone outside for carrying out the morning observation later 445
and later. This might explain the trend shift in the morning observation. Following this 446
hypothesis the minimum temperature was adjusted, ΔT, by use of formula (2) for the winter 447
season in accordance with the regression line shown in Fig. 7, where a = year (period 1832-448
1838). No adjustments were undertaken for the period 1829-1831. 449
450
∆𝑇=0.2861 ∙𝑎−523.85 (2) 451
452
4.4 Overheating of the midday observation 453
The midday observation turned out to be homogenous, but it might have been overheated by 454
insufficient radiation protection in Esmark’s yard. This was tested by comparison with the 455
Oslo – Blindern station that is well protected by a Stevenson screen. Difference between the 456
midday observation and the evening observation reveals a characteristic pattern (Fig. 8). 457
Whereas the differences were almost equal in the months September – March, the differences 458
in the Esmark series were larger than the differences in the Blindern series for the months 459
April – August. They were particularly large in MJA where the sun is highest on the sky and 460
the radiation reaches its annual maximum. Therefore our interpretation is that Esmark’s 461
thermometer was overheated at the midday observation by (reflected) short wave radiation in 462
the period April – August, but not for the rest of the year. Based on the differences between 463
the two curves the adjustments of the midday observation are also given (lower panel in Fig. 464
8). 465
466
4.5 Homogenisation of the monthly mean temperature. 467
Esmark observed only three times a day, so it is far from obvious how monthly mean 468
temperature should be calculated without bias. This problem confronts meteorological 469
institutes worldwide so formulas for the calculation are developed (see Appendix B). The 470
Clim. Past Discuss., doi:10.5194/cp-2016-60, 2016
Manuscript under review for journal Clim. Past
Published: 20 June 2016
c
Author(s) 2016. CC-BY 3.0 License.
15
formulas contain specific constants valid for each month and site. Strictly speaking the 471
constants were unknown for Esmark’s observation site at Øvre Vollgate, but well known for 472
the station 18700 Oslo – Blindern lying 3.4 km to the north of Esmark’s site. Fortunately 473
there are indications that the constants for Blindern could be used also for Øvre Vollgate (see 474
Appendix 2). Given the constants the calculation of homogenous monthly mean temperature 475
was trivial when the homogenised version of the observations at fixed hours was used. We 476
found that the adjustments for seasonal means vary from -0.7ᵒC to +0.3ᵒC (Fig. 9). The 477
adjustments were negative except from the last part of the series in winter and autumn. For the 478
annual means the adjustments are much less, they vary from -0.4ᵒC to -0.1ᵒC. 479
4.6 The climate in Esmark’s period of observation, 1816-1838 480
Esmark’s observations exhibit a long-term variation pattern characterised by lower values in 481
the start and in the end of the period, whereas the middle of the period was somewhat warmer, 482
cf. Fig. 10. This is true not only for the annual means, but also for all seasons of the year. For 483
individual years 1822 is warmest except in summer. The coldest year is 1816 followed by the 484
years 1817, 1820 and the last one 1838. In the year 1816 stands out as coldest also in two 485
seasons, spring (MAM) and autumn (SON), and also in the two individual months March and 486
May (not shown). 487
The year 1816 is of particular interest as it has gone into history as “the year without 488
summer” (Fagan, 2001). However, Esmark’s observations show that this summer (JJA) was 489
not very extraordinary in Oslo, as the following summer of 1817 was colder, and in particular 490
that of 1821. More extraordinary is the spring temperature in 1816, being the only one with 491
mean temperature below zero. For agriculture the first years of Esmark’s period of 492
observation must have been bad taking into account that low temperature is a limiting factor. 493
For the grain growing months (AMJJA) the mean temperature was about 8.5ᵒC in the three 494
consecutive years 1816, 1817 and 1818, i.e. the lowest temperatures in Esmark series of 495
observation. 496
5"Discussion"497
From 1816 to the mid-1820s the annual Christiania temperature as recorded by Esmark rose 498
by approximately 1.5oC, then subsequenly slowly fell by almost 1oC towards 1840 (Fig. 10). 499
This general pattern is consistent with that found for the same time interval in the Swedish 500
capital Stockholm (compare with Fig. 5 in Moberg et al., 2002). 501
502
Clim. Past Discuss., doi:10.5194/cp-2016-60, 2016
Manuscript under review for journal Clim. Past
Published: 20 June 2016
c
Author(s) 2016. CC-BY 3.0 License.
16
5.1 Adjusting for inhomogeneities 503
An important inhomogeneity was detected in Esmark’s data at the end of 1822 in the evening 504
observation, and was adjusted for. Alternatively the inhomogeneity could be considered only 505
as a change of observational time, and not adjusted for by the testing. The series of mean 506
temperatures could then have been kept homogenous by assessing how much the observation 507
time had changed, leading to a corresponding change in the constants in Føyn’s formula for 508
calculation of monthly mean temperature (see Appendix B). Probably also other changes 509
could have taken place at the end of 1822, so therefore we considered it better to apply the 510
adjustments directly to the temperature data, and use the same constants on both sides of the 511
shift for mean monthly temperature calculation. Moreover, there is some indication that a 512
changed environment could have played a role for this inhomogeneity as Esmark in 1823 513
planted fruit trees in his garden, cf. Metadata. 514
No doubt Esmark possessed a minimum thermometer from 1828. Such instruments were 515
available even before Esmark started his Oslo series in 1816. Already in 1790 a spirit 516
thermometer with a glass index, very much like those used up to this day at manual stations, 517
was described to the Royal Society in Edinburgh (Middleton, 1966: p. 152). In our work the 518
change from an ordinary thermometer reading at the morning observation to a minimum 519
thermometer reading was accounted for by a change of formula for mean monthly 520
temperature calculation. Therefore the series of mean monthly temperatures was kept 521
homogenous without adjusting for this shift in the morning observation. 522
The size of the adjustments of Esmark’s observations gives an indication of the uncertainty of 523
Esmark’s observations ( Fig. 9). These are adjustments for both homogeneity errors and short 524
wave radiation errors. They are largest during summer, which also are expected due to the 525
lack of radiation screens other than the wall of houses. For annual mean temperature the 526
adjustments are within the interval [-0.4ᵒC, -0.1ᵒC]. For individual observation times the 527
adjustments were higher [-0.7ᵒC, +0.3ᵒC]. 528
529
5.2 Comparison with other observations 530
During the period 1822.11-1827.02 the Christiania professor Christopher Hansteen carried 531
out observations at his home in Pilestredet at the corner of Keysersgate (Hansteen 1823, 1824, 532
1828; Birkeland, 1926: p. 12). The distance from Esmark’s site was only about 600 m. 533
Hansteen’s observation times varied much but for each month he gives the observation times 534
together with the data (Hansteen, 1824). The distribution of the observation times in UTC is 535
as follows: morning 06h 4%, 07h 44%, 08h 52%; midday 13h 20%, 14h 78%, 15h 2%; evening 536
Clim. Past Discuss., doi:10.5194/cp-2016-60, 2016
Manuscript under review for journal Clim. Past
Published: 20 June 2016
c
Author(s) 2016. CC-BY 3.0 License.
17
21h 6%, 22h 88%, 23h 6%. Hansteen’s observations were adjusted to Esmark’s observation 537
times, approximately 08, 15 and 21 UTC by use of the mean daily temperature wave at 538
Blindern so that Esmark’s observations could be compared with the adjusted ones of 539
Hansteen, Fig 11. It is evident that Hansteen’s morning observation is much warmer than that 540
of Esmark except during winter. Much likely the thermometers of Hansteen had been 541
overheated as his two thermometers hang at the southern and northern side of the house 542
(Birkeland, 1925: 12). Then it must have been difficult to find shadow in the morning. Also 543
the midday observation is warmer by Hansteen than by Esmark. This is harder to understand. 544
If Birkeland’s account of the thermometer at the north wall of the house is correct the house is 545
expected to give sufficient protection of that thermometer (Nordli et al., 2015), but as nothing 546
is known about the environment other factors might have been involved. 547
The evening temperature, however, is much in agreement with that of Esmark during summer 548
unlike for the two other observation times. The evening observations occurred after sunset at 549
both sites, whereas the two other observations occurred after sunrise. This supports the 550
suggestion that the differences at the morning and midday observations are due to radiation 551
errors. 552
553
Unlike the situation during summer, Hansteen’s temperatures are lower than those of Esmark 554
in the period November – March (Fig. 11). In many weather situations the air loses energy by 555
long wave radiation because the short wave radiation is too small to compensate for the loss. 556
The result is that the coldest air is found at the lowest places in the local terrain, not 557
necessarily at the lowest sites above sea level. Esmark’s house lies high in the local terrain at 558
the edge of a slope down to Pipervika cf. Metadata, whereas Hansteen’s house lies low in the 559
local terrain at a floor of a small valley. The difference in winter temperature is therefore must 560
likely due to different local climate. 561
562
At The Astronomical Observatory in Oslo meteorological observations started in April 1837 563
that lasted almost for one hundred years (Nordli et al., 2015), so this series overlaps Esmark’s 564
series by 21 months. For comparison of the two series we have used unadjusted observations 565
from the observatory, whereas both adjusted and unadjusted Esmark observations are used 566
(Fig. 12). It is evident that for all seasons but winter Esmark’s temperatures are lower than 567
those from the Observatory. Esmark died on 26 January 1839 (see Metadata), so probably the 568
quality of the latest months of his series may be questioned. However, we cannot see any 569
declined quality directly from is observation protocols, but it is possible that the last two years 570
Clim. Past Discuss., doi:10.5194/cp-2016-60, 2016
Manuscript under review for journal Clim. Past
Published: 20 June 2016
c
Author(s) 2016. CC-BY 3.0 License.
18
of his observations are not representative for Esmark’s observational practice. Moreover, the 571
overlapping period is very short; only two years for most of the months, and only one year for 572
the months January to March. It is therefore possible that the present comparison is not valid 573
for Esmark’s entire period of observation. 574
575
5.3 The accuracy of the thermometers 576
In Esmark’s protocol for 1816 some instrumental corrections are given for what is claimed to 577
be Esmark’s thermometer, Table 6. They are not written by Esmark himself, most probably 578
they are notes written by Birkeland, who says he has them after Hansteen 1821-23, but it is 579
not certain that they belong to the thermometer used by Esmark. The corrections are very 580
small for the frequent winter temperatures, but as high as 0.5ᵒC for frequent summer 581
temperatures. Due to the uncertainty with the identification of Esmark’s thermometer we have 582
not applied the corrections to his observations. It should also be kept in mind that Esmark 583
used another thermometer, i.e. a minimum thermometer for the period 1828.03-1838.12, 584
which might also have instrumental corrections. However, Esmark was a skilled instrument 585
builder, so it is not likely that he used thermometer with larger corrections that those in Table 586
6. 587
588
There were several volcanic eruptions that affected the world climate in the first years of 589
Esmark’s period of observation. The Tambora eruption in 1815 was probably the greatest one. 590
It has given rise to the paradigm for 1816: “the year without a summer”. Esmark’s 591
observations show, however, that albeit being cold the summer was not extraordinary cold in 592
Oslo. And in the Stockholm series (“Bolin Centre Database,”) the summer of 1816 was rather 593
warm, No 17 of the 23 summers from 1816-1838, ranged from low to high (Table 7). May, 594
however, was very cold in both cities, and July was quite warm in both cities, but in June and 595
August Oslo was much colder relative to the mean value than Stockholm. 596
597
There exist climate reconstructions for the period 1816-1838, independent from Esmark’s 598
observations, based upon ice loss from Lake Randsfjorden (Nordli et al., 2007) temperature 599
proxy for the season February-April, and upon the date of grain harvest for Austlandet 600
(Nordli, 2001a), Vestlandet (Nordli et al., 2003), Lesja (Nordli, 2001b) and Trøndelag 601
(Nordli, 2004) temperature proxies for the seasons April-August and May-August (Table 8). 602
The three reconstructions within the county of South-Eastern Norway are all in agreement 603
with Esmark’s observations that the summer of 1816 was among the coldest in the grain 604
Clim. Past Discuss., doi:10.5194/cp-2016-60, 2016
Manuscript under review for journal Clim. Past
Published: 20 June 2016
c
Author(s) 2016. CC-BY 3.0 License.
19
growing seasons, whereas the reconstructions for the two other counties, Western and Mid 605
Norway, show relatively warm summers, even more so than those in Stockholm. 606
607
Anomalies of surface temperature and precipitation for the summer months of 1816 has been 608
reconstructed (Luterbacher and Pfister, 2015). They show a positive gradient from a cold core 609
of air lying over France towards Eastern and Northern Europe, so the paradigm of the severe 610
summer of 1816 has to be modified when it comes to Scandinavia and Eastern Europe. It 611
looks like this is easy to forget, e.g. “…weather patterns were disrupted worldwide for 612
months, allowing for excessive rain, frost, and snowfall through much of the Northeastern 613
U.S. and Europe in the summer of 1816”(Klingaman and Klingaman, 2014). It is therefore 614
important that the temperature gradient is recognised. The results in Table 8 are a part of the 615
pattern showing the spatial variability in Europe that summer. 616
6"Conclusions"617
Esmark’s observations are almost complete for the years 1816-1838. Homogeneity testing 618
revealed a shift in the evening observation at the end of 1822. From March 1828 Esmark 619
noted nightly minimum temperature instead his previous notation of morning temperature. 620
During the years 1831 to 1838 the nightly minimum temperature increased almost steadily in 621
the winter season, i.e. it was inhomogenous. The homogenized temperature series showed low 622
temperature in both ends of the series, with higher temperature in the middle, i.s. the 1820s. 623
The starting year, 1816, is of particular interest as it has been referred to as the year without a 624
summer. The summer in Oslo was cold, but not extraordinary cold, as it was only the third 625
coldest in the period of observation. However, the annual mean of 1816 and also the months 626
March and May that year were the coldest ones in that period. 627
628
629
630
631
632
633
Clim. Past Discuss., doi:10.5194/cp-2016-60, 2016
Manuscript under review for journal Clim. Past
Published: 20 June 2016
c
Author(s) 2016. CC-BY 3.0 License.
20
REFERENCES 634
Alexandersson, H., 1986. A homogeneity test applied to precipitation data. J. Climatol. 6, 635
661–675. doi:10.1002/joc.3370060607 636
Alexandersson, H., Moberg, A., 1997. Homogenization of Swedish Temperature Data. Part I: 637
Homogeneity Test for Linear Trends. Int. J. Climatol. 17, 25–34. 638
doi:10.1002/(SICI)1097-0088(199701)17:1<25::AID-JOC103>3.0.CO;2-J 639
Andersen, B. G. 1992. Jens Esmark – a pioneer in glacial geology. Boreas 21: 97-640
102. 641
Anonymous 1839. Biografi öfver Jens Esmark, Professor i Bergvetenskapen vid Universitetet 642
i Christiania, Riddar af Kongl. Wasa-Orden. Kgl. [svenska] Vetenskaps-Akademien, 643
Nya Handlingar 1838: 312-323. [The biography was written by Esmark’s son Hans 644
Morten Thrane Esmark & Jøns Jacob Berzelius] 645
Auer, I., Böhm, R., Jurkovic, A., Lipa, W., Orlik, A., Potzmann, R., Schöner, W., 646
Ungersböck, M., Matulla, C., Briffa, K., Jones, P., Efthymiadis, D., Brunetti, M., 647
Nanni, T., Maugeri, M., Mercalli, L., Mestre, O., Moisselin, J.-M., Begert, M., Müller-648
Westermeier, G., Kveton, V., Bochnicek, O., Stastny, P., Lapin, M., Szalai, S., 649
Szentimrey, T., Cegnar, T., Dolinar, M., Gajic-Capka, M., Zaninovic, K., Majstorovic, 650
Z., Nieplova, E., 2007. HISTALP—historical instrumental climatological surface time 651
series of the Greater Alpine Region. Int. J. Climatol. 27, 17–46. doi:10.1002/joc.1377 652
Berger, A., 2012. A Brief History of the Astronomical Theories of Paleoclimates. Pp. 107-653
129. In : A. Berger (ed.) Climate Change. Wien : Springer Verlag. 10.1007/978-3-654
7091-0973-1_8 655
Moberg, A., Bergström, H., Krigsman, J. R. & Svanered, O. 2002. Daily air temperature and 656
pressure series for Stockholm (1756-1998). Climatic Change 53: 171-212. 657
Bergström, H., Moberg, A., 2002. Daily Air Temperature and Pressure Series for Uppsala (1722–658
1998). Clim. Change 53, 213–252. doi:10.1023/A:1014983229213. 659
Birkeland, B. J., 1925. Ältere Meteorologische Beobachtungen in Oslo (Kristiania). Luftdruck 660
und Temperatur in 100 Jahren. (Old meteorological observations in Oslo. One hundred 661
years of air pressure and temperature). (submitted 1923). Geofys. Publ. III, 56 pp. 662
Bolin Centre Database [WWW Document], n.d. URL 663
http://bolin.su.se/data/stockholm/homogenized_monthly_mean_temperatures.php 664
(accessed 2.10.16). 665
Document 1. Riksarkivet (National archive), Oslo, Christiania Byfogden A 666
Tinglysning, Tinglysninger frem til 1819, property No. 308. Also: Kristiania 667
Clim. Past Discuss., doi:10.5194/cp-2016-60, 2016
Manuscript under review for journal Clim. Past
Published: 20 June 2016
c
Author(s) 2016. CC-BY 3.0 License.
21
Magistrat Diverse Pakkesaker Branntakster L – Lac 0009 poliser No. 308. 668
Protocol No. 333, Grundtaxt No. 308 i Øvre voldgade. 669
Document 2. Riksarkivet (National archive), Oslo, B VII 1 Kristiania Magistrat Fa 670
– Folketellinger 0001 (1815); Ga – Manntall 0004 (1816), 0005 (1826, 671
1828), 0006 (1829, 1830), 0007 (1831, 1833), 0008 (1834), 0009 (1835, 672
1836), 0010 (1837); Kristiania Magistrat skatter Gc. 673
Document 3. Riksarkivet (National archive), Oslo, B VII 1 Kristiania Magistrat Fa 674
– Folketellinger 0001 (1815) April-May 1815, list of inhabitants in Søndre 675
Rode, No. 226; and Christiania Mandtallsregister 26 June 1815, grunn og 676
næring. No. 221. See also Riksarkivet, Oslo, Christiania Byfogden A 677
Tinglysning, Tinglysninger frem til 1819. No. 221. 678
Document 4. Oslo Byarkiv (City archive), Christiania matrikuleringsprotokoll, 679
1830, page 142. Document available as PDF at Oslo Kommune, Plan og 680
bygningsetaten. 681
Document 5. Oslo Kommune (Oslo municipality), Plan og bygningsetaten, 1938, 682
Journalnummer 1768/1938 Riving av (demolition of) Øvre Vollgt. 7. 683
Document copy available on fiche. 684
Document 6. Oslo Byarkiv (City archive), Christiania matrikuleringsprotokoll, 685
1830, pp. 141, 142, 143 & 163. Documents available as PDF at Oslo 686
Kommune, Plan og bygningsetaten. 687
Document 7. Landarkivet (County archive), Fyn, Denmark. Stamhuset 688
Hofmansgaves Arkiv. J. Esmark, letter to Nils Hofman Bang 31 October 689
1823. Thanks for fruit trees which are now all planted in his garden. 690
Ducré-Robitaille, J.-F., Vincent, L.A., Boulet, G., 2003. Comparison of techniques 691
for detection of discontinuities in temperature series. Int. J. Climatol. 23, 692
1087–1101. doi:10.1002/joc.924 693
Esmark, J. 1799-1810. Esmark’s handwritten observation protocols from 694
Kongsberg. 11 volumes (1802 missing). Riksarkivet (National archive), 695
Oslo. S-1570. Det norske meteorologiske institutt. F/Fa. Materiale etter 696
professorer. L0001. Esmark’s Kongsberg protocols. 697
Esmark, J. ”1802”. [1803] Auszug aus einem Schreiben des Oberbergamts-698
Assessor J. Esmarck zu Kongsberg über die Schnee- und Vegetations-linie in 699
Norwegen. Nordisches Archiv für Natur- und Artzneywissenschaft und 700
Chirurgie III (3): 197-200. Copenhagen. 701
Clim. Past Discuss., doi:10.5194/cp-2016-60, 2016
Manuscript under review for journal Clim. Past
Published: 20 June 2016
c
Author(s) 2016. CC-BY 3.0 License.
22
Esmark, J. 1812. Bemærkninger, gjorte paa en Reise til Gousta-Fjeldet i Øvre-702
Tellemarken. Dated Kongsberg 29 December 1810. Topographisk-Statistiske 703
Samlinger. Udgivne af Det Kongelige Selskab for Norges Vel. Første Deels 704
Andet Bind. Christiania. Trykt hos W. Wulfsberg. pp. 175-196. 705
Esmark, J. 1816-1838. Esmark’s original handwritten observation protocols from 706
Christiania. 23 volumes. Riksarkivet (National archives), Oslo. S-1570. Det 707
norske meteorologiske institutt. F/Fa. Materiale etter professorer. L0002. 708
Esmark’s Christiania protocols. 709
Esmark, J. 1820. Et nyt Hygrometer. Den Norske Rigstidende, No. 84 (20 710
October). 711
Esmark, J. 1823. Middel-Barometerstand og Middel-Temperatur for Christiania i 712
de syv Aar fra 1816 til 1822. Magazin for Naturvidenskaberne. (Förste 713
Aargangs förste Bind) [1]: [p. 178 – unpaginated table at end of volume]. 714
Esmark, J. 1824. Bidrag til vor Jordklodes Historie. Magazin for 715
Naturvidenskaberne. (Anden Aargangs förste Bind, Förste Hefte) [3]: 28-49. 716
Esmark, J. 1825. Handwritten eight-page vitae/autobiography, Christiania 15 717
October 1825. Kungliga Vetenskapsakademien - Royal Swedish Academy of 718
Science, Stockholm, Center for the History of Science, Archives, category 719
“Inkommande skrivelser från personer utan eget arkiv”, 720
Esmark, J. 1826 Remarks tending to explain the Geological History of the Earth. 721
The Edinburgh New Philosophical Journal 2: 107-121. 722
Esmark, J. 1833. Thermometer- og Barometer-Stand i Christiania efter 16325 723
Observationer i 15 år. Eyr: et medicinsk Tidsskrift 8 : 235-239. Christiania. 724
Fagan, B., 2001. The Little Ice Age: How Climate Made History 1300-1850, 1 725
edition. ed. Basic Books. 726
[Fearnley, C.] 1865. Meteorologische Beobachtungen an der Königlichen Universitäts-727
Sternwarte zu Christiania. 1837-1863. Christiania: H. J. Jensen. 728
Gjelten, H.M., Nordli, Ø., Grimenes, A.A., Lundstad, E., 2014. The Ås Temperature Series in 729
Southern Norway–Homogeneity Testing And Climate Analysis. Bull. Geogr. Phys. 730
Geogr. Ser. 7, 7–26. doi:10.2478/bgeo-2014-0001 731
Hansteen, C. 1823. Meteorologisk Dagbog for den sidste Fjerdedeel af 1822. 732
Magazin for Naturvidenskaberne. Förste Aargangs förste Bind, 1. Hefte, [p. 733
177 – unpaginated table at end of volume]. 734
Clim. Past Discuss., doi:10.5194/cp-2016-60, 2016
Manuscript under review for journal Clim. Past
Published: 20 June 2016
c
Author(s) 2016. CC-BY 3.0 License.
23
Hansteen, C. 1824. Forelöbige Resultater af Barometer-Iagttagelser i Christiania. 735
Magazin for Naturvidenskaberne. (Anden Aargangs 2 Hefte) [4]: 269-298. 736
Hansteen, Christopher 1828. Timevise Thermometer- og Barometer-Iagttagelser i 737
Trondhjem. Magazin for Naturvidenskaberne 8 (1): 173. 738
Hansteen, C. 1841. Resultaterne af tre Aars Barometer-Iagttagelser i Christiania. 739
In: Forhandlinger ved de skandinaviske Naturforskeres andet Möde, der 740
holdtes i Kjöbenhavn fra den 3die til den 9de Juli 1840. Kjöbenhavn: I 741
Commission hos Universitetsboghandler C.A. Reitzel. pp. 52-64. 742
Hestmark, G. 2009. ”Her ligger Sneen evig.” Da Dovre falt – for Esmarks 743
barometer. Historisk Tidsskrift 88: 231-249. 744
Horrebow, P. 1780. Tractatus historico-meteorologicus. Havniæ. 745
Klingaman, W.K., Klingaman, N.P., 2014. The Year Without Summer: 1816 and 746
the Volcano That Darkened the World and Changed History, Reprint edition. 747
ed. St. Martin’s Griffin. 748
Kratzenstein, C. G. 1791. Forelæsninger over Experimental-Physiken. 749
Kiöbenhavn: Trygt hos Johan Frederik Schultz. Hos Faber og Nitsche. 750
Krüger, T. 2013. Discovering the Ice Ages: International Reception and 751
Consequences for a Historical Understanding of Climate. (History and 752
Medicine Library 37). Leiden: Brill, 2013. 753
Luterbacher, J. & Pfister, C. The year without summer. Nature Geoscience 8: 246-754
248. 755
Livingstone, J. 1819. Account of an improved Hygrometer. The Edinburgh 756
Philosophical Journal 1: 116-117. 757
Myhre, J. E. 1990. Oslo Bys Historie. Vol. 3. Hovedstaden Christiania. Fra 1814 758
til 1900 . Oslo: 759
Moberg, A., Alexandersson, H., 1997. Homogenization of Swedish Temperature Data. Part II: 760
Homogenized Gridded Air Temperature Compared with a Subset of Global Gridded Air 761
Temperature Since 1861. Int. J. Climatol. 17, 35–54. doi:10.1002/(SICI)1097-762
0088(199701)17:1<35::AID-JOC104>3.0.CO;2-F 763
Nordli, 2001a. Spring and summer temperatures in south eastern Norway (1749 – 2000) 764
(DNMI-klima No. 01/2001). 765
Nordli, Ø., 2004. Spring and summer temperatures in Trøndelag 1701 – 2003 (met.no/report 766
No. 05/2004). Meteorological Institute, Oslo. 767
Clim. Past Discuss., doi:10.5194/cp-2016-60, 2016
Manuscript under review for journal Clim. Past
Published: 20 June 2016
c
Author(s) 2016. CC-BY 3.0 License.
24
Nordli, Ø., Hestmark, G., Benestad, R.E., Isaksen, K., 2015. The Oslo temperature series 768
1837–2012: homogeneity testing and temperature analysis. Int. J. Climatol. 35, 3486–769
3504. doi:10.1002/joc.4223 770
Nordli, Ø., Lie, Ø., Nesje, A., Dahl, S.O., 2003. Spring–summer temperature reconstruction in 771
western Norway 1734–2003: a data-synthesis approach. Int. J. Climatol. 23, 1821–1841. 772
doi:10.1002/joc.980 773
Nordli, Ø., Lundstaf, E., Ogilvie, A.E.J., 2007. A Late Winter-Early Spring Temperature 774
Reconstruction for Southeastern Norway from 1758 to 2006. ??? 46, 404–408. 775
Nordli, P.Ø., 2001b. Reconstruction of Nineteenth Century Summer Temperatures in Norway 776
by Proxy Data from Farmers’Diaries. Clim. Change 48, 201–218. 777
doi:10.1023/A:1005698302572 778
Rudwick, M. J. S. 2008. Worlds Before Adam. The Reconstruction of Geohistory in the Age 779
of Reform. Chicago & London: The University of Chicago Press. 780
Snorrason, E. 1974. C. G. Kratzenstein: professor physices experimentalis Petropol. et Havn. 781
and his Studies on electricity during the eighteenth century. Acta historica scientiarum 782
naturalium et medicinalium no. 29 / edidit Bibliotheca Universitatis Hauniensis. Odense: 783
Odense University Press. 784
Splinter, Susan 2007. Zwischen Nützlichkeit und Nachahmung. Eine Biografie des Gelehrten 785
Christian Gottlieb Kratzenstein (1723–1795). Frankfurt (Main): Peter Lang. 786
Tuomenvirta, H., 2001. Homogeneity adjustments of temperature and precipitation series—787
Finnish and Nordic data. Int. J. Climatol. 21, 495–506. doi:10.1002/joc.616 788
Willaume-Jantzen, V. 1896. Meteorologiske Observationer i Kjøbenhavn. Résumé des 789
Observations Météorologiques de Copenhague. Det Danske Meteorologiske Institut. 790
Kjøbenhavn, i commission hos universitets-boghandler G.E.C. Gad. 791
Worsley, P. 2006. Jens Esmark, Vassryggen and early glacial theory in Britain. Mercian 792
Geologist 16: 161-172. 793
794
Clim. Past Discuss., doi:10.5194/cp-2016-60, 2016
Manuscript under review for journal Clim. Past
Published: 20 June 2016
c
Author(s) 2016. CC-BY 3.0 License.
25
795
Figure texts 796
797
Fig. 1. Map of Christiania (now Oslo) 1811 with the location (red star) of 798
Esmark’s house in Øvre Vollgt. 7 marked. 799
800
Fig. 2. Matriculation and survey 1830 of Esmark’s property No. 308, Øvre Voldgate 7, in 801
Oslo Byarkiv (City archives). Arrow indicates N. Garden to the left, house surrsounding back 802
yard to the right. 803
804
Fig. 3. Street view of Esmark’s house in Øvre Voldgate 7. Photograph from around 1900. 805
Oslo Bymuseum, No. OB.F00897. High buildings on each side built late 19th century. 806
807
Fig. 4. Esmark’s Christiania protocol for 1817. Now deposited at Riksarkivet 808
(National archives), Oslo. S-1570. Det norske meteorologiske institutt. F/Fa. 809
Materiale etter professorer. L0002. 810
811
Fig. 5. The January page from Esmark’s meteorological observation protocol from 812
1823, the year he discovered ice ages. Now deposited at Riksarkivet (National 813
archives), Oslo. S-1570. Det norske meteorologiske institutt. F/Fa. Materiale etter 814
professorer. L0002. 815
816
Fig. 6. The first published Christiania weather table, from Den norske Rigstidende, 817
24 January 1818. 818
819
Fig. 7 The temperature difference (ᵒC) between Esmark’s evening observation and the 820
morning observation the following day for the winter season (Dec-Feb). 821
822
Fig. 8 Temperature differences (ᵒC) between the observations at Blindern at 15 UTC and at 21 823
UTC for the period 1993.01-2015.09. Also the difference between the midday and evening 824
observations of Esmark is shown for the period 1816.01-1838.12. (The adjustments of the 825
evening observations, Table 5, are added to the data for the period 1816.01-1821.12 before 826
the calculation of the differences. In the table below the figure are shown the adjustments of 827
Esmark’s midday observation 828
829
Fig. 9. Adjustments added to Esmark’s series for each season during his period of 830
observation, 1816-1838. 831
832
Fig. 10. Annual and seasonal means of Esmark’s temperature series (symbols), and Gaussian 833
filter (curves) with standard deviation 3 in the Gaussian distribution (e.g. Nordli et al., 2015), 834
corresponsing roughly to a 10 year regtangular filter. 835
836
Clim. Past Discuss., doi:10.5194/cp-2016-60, 2016
Manuscript under review for journal Clim. Past
Published: 20 June 2016
c
Author(s) 2016. CC-BY 3.0 License.
26
Fig. 11. Difference between Esmark’s observations at Øvre Vollgate and Hansteen’s 837
observations at Pilestredet (Esmark minus Hansteen) during the period 1822.11-1827.02 at 838
08, 15 and 21 UTC. The monthly means are calculated by Føyn’s formula, cf. Appendix 1 839
840
Fig. 12. Differences in mean monthly temperature between Esmark’s observations at Øvre 841
Vollgate and those at the Astronomical Observatory (Esmark minus Observatory) during the 842
period 1837.04-1838.12. Esmark’s observations are presented both unadjusted and adjusted. 843
For the observatory the temperatures are unadjusted. 844
845
846
847
848
849
850
851
Clim. Past Discuss., doi:10.5194/cp-2016-60, 2016
Manuscript under review for journal Clim. Past
Published: 20 June 2016
c
Author(s) 2016. CC-BY 3.0 License.
27
Tables 852
853
Table 1 Esmark’s station at Øvre Vollgate 7 as well as other observation stations used in this article: 854
national station number (identifier) and name, period of observation and station altitude. The star 855
before the start year marks the start of hourly observations 856
857
No. and name
Period (from-to; year, month, day
Hs (m)
18651 Oslo II
1837.04.02-1933.12.31
25
18654 Oslo - Øvre Vollgate
1816.01.01-1838.12.31
11
18655 Oslo - Pilestredet
1822.10.19-1827.02.28
16
18700 Oslo - Blindern
*1993.01.05 to present
94
18815 Oslo - Bygdøy
*2012.01.01 to present
15
858
859
Table 2 The SNHT test used for comparison of observations at time x versus observations at time y (x 860
vs y). The shifts are given by the last year of each part of the series. For the single shift test also the 861
adjustment needed for the x-series to be homogenous with y-series (Non-significant results are given 862
in italic). 863
Part 1, 1816.01-1838.12: The whole length of the series
SNHT tests
Obs. times
Winter
Spring
Summer
Autumn
Year
Single shift
I vs II
1833; -1.1
1827; -2.1
1827; -3.3
1824; -1.4
1827; -1.8
Single shift
I vs III
1832; -1.5
1826; -2.8
1827; -4.0
1827; -1.7
1827; -2.4
Single shift
III vs II
1821; 0.7
1820; 1.5
1821; 1.3
1821; 0.6
1821; 0.9
Double shift
I vs II
1826; 1834
1818; 1827
1817; 1827
1824; 1829
1823; 1827
Double shift
I vs III
1819; 1832
1820; 1826
1818; 1828
1823; 1829
1818; 1827
Double shift
III vs II
1821; 1832
1819; 1835
1821; 1835
1817; 1834
1821; 1835
Part 2, 1816.01 – 1828.02
SNHT-tests
Obs. times
Winter
Spring
Summer
Autumn
Year
Single shift
I / II
1826; -0.8
1818; -0.7
1817; -0.8
1824; -1.0
1823; -0.5
Single shift
I /III
1818; -1.0
1820; -1.7
1818; -1.7
1821; -0.9
1818; -1.3
Single shift
III / II
1821; 0.6
1819; 1.4
1821; 1.2
1817; 0.8
1821; 0.8
Part 3, 1828.03 – 1838.12
SNHT-tests
Obs. times
Winter
Spring
Summer
Autumn
Year
Single shift
I / II
1834; -1.0
1834; 0.4
1830; -0.4
1829; -0.4
1830; -0.5
Single shift
I /III
1832; -1.3
1836; -0.6
1836; -0.8
1829; -0.9
1836; -0.8
Single shift
III / II
1833; 0.4
1835; 0.8
1835; 0.9
1834; 0.6
1835; 0.7
864
865
Clim. Past Discuss., doi:10.5194/cp-2016-60, 2016
Manuscript under review for journal Clim. Past
Published: 20 June 2016
c
Author(s) 2016. CC-BY 3.0 License.
28
Table 3. The same as Table 1, but the single shift test used on monthly resolution. In the 1st and 3rd 866
rows are given the year of the shifts, and in the 2nd and 4th rows the adjustments. 867
Jan
Feb
Mar
Apr
May
Jun
Jul
Aug
Sep
Oct
Nov
Dec
I/II
1834
-1.2
1826
-1.4
1826
-1.0
1830
-2.2
1827
-3.3
1827
-3.4
1827
-3.5
1827
-2.9
1825
-1.9
1827
-1.1
1824
-1.5
1833
-1.2
III/II
1828
0.6
1832
0.7
1820
1.1
1819
1.7
1819
1.8
1826
1.3
1821
1.3
1821
1.3
1821
0.8
1820
0.9
1834
0.6
1820
0.7
868
Table 4 Difference, Diff (ᵒC), of median temperature between Esmark’s evening observations and the 869
observations the following morning during different time intervals. The similar differences for the 870
modern station Oslo – Blindern are also shown, i.e. the observation at 21 UTC and the minimum 871
temperature at 08 UTC. Also the standard deviations, STD (ᵒC), of the differences are shown. 872
Jan
Feb
Mar
Apr
May
Jun
Jul
Aug
Sep
Oct
Nov
Dec
Esmark
1816.01-1821.12
Diff
STD
0.0
3.4
0.0
2.6
0.0
2.4
-0.7
2.1
-1.8
2.4
-1.6
2.3
-1.3
2.6
-1.2
2.1
0.0
2.1
0.5
2.0
0.0
2.6
0.0
2.2
Esmark
1822.01-1828.02
Diff
STD
0.9
3.1
0.7
2.5
1.2
2.3
0.6
1.8
0.6
2.2
-0.7
2.4
-0.6
2.2
0.0
2.1
1.2
2.9
0.6
2.5
0.8
2.5
0.6
2.4
Esmark
1828.03-1838.12
Diff
STD
1.3
2.6
1.5
2.3
1.9
2.5
2.2
1.8
3.1
2.1
3.1
2.2
3.1
2.4
3.1
2.3
2.5
2.2
1.9
2.1
1.6
1.9
1.3
2.7
Blindern
1993.09-2015.09
Diff
STD
1.0
1.7
1.5
1.8
2.3
1.8
2.6
1.7
3.2
1.8
3.0
1.8
2.7
1.7
2.4
1.6
2.0
1.6
1.5
1.6
1.0
1.5
1.0
1.6
873
Table 5 Adjustment (ᵒC) of the evening observation in the period 1816.01-1821.12 874
!"#$
%&'$
(")$
*+)$
(",$
!-#$
!-.$
*-/$
0&+$
123$
456$
7&2$
89:;$
89:;$
8<:9$
8<:=$
8<:>$
8<:=$
8<:>$
8<:>$
89:?$
89:@$
89:>$
89:A$
875
Table 6. Instrument correction (Corr) for thermometer readings (Temp.). The thermometer might have 876
been used by Esmark, 1816-1838. 877
Temp. (ᵒC)
25.00
18.75
12.50
6.25
0.00
-6.25
-12.50
-18.75
-25.00
Corr. (ᵒC)
+0.50
+0.50
+0.38
+0.38
+0.13
+0.13
+0.13
+0.13
+0.63
878
Table 7 The range of mean temperature in 1816 for months and seasons during the years 1816-1838 879
for Oslo (Esmark’s observations). For comparison also Stockholm is included. The range runs from 880
low to high values. 881
J
F
M
A
M
J
J
A
S
O
N
D
Yr
Wi
Sp
Su
Au
Oslo
14
5
1
3
1
3
12
2
2
3
7
10
1
9
1
3
2
Stockholm
14
3
6
9
1
16
18
9
13
5
8
12
7
6
4
17
3
882
Clim. Past Discuss., doi:10.5194/cp-2016-60, 2016
Manuscript under review for journal Clim. Past
Published: 20 June 2016
c
Author(s) 2016. CC-BY 3.0 License.
29
Table 8 The range of mean temperature in 1816 for seasons during the years 1816-1838 for Oslo 883
(Esmark’s observations), and for climate reconstructions from proxy data at different places in Norway. 884
For comparison also Stockholm is included. The range runs from low to high values. 885
Place, County
February - April
April – August
May-August
Oslo, South-eastern Norway
2
1
3
Randsfjorden, South-eastern Norway
2
Austlandet, South Eastern Norway
1
Lesja, South-eastern Norway
1
Bergen, Western Norway
18
Trøndelag, Mid Norway
18
Stockholm, Sweden
3
10
9
886
887
Clim. Past Discuss., doi:10.5194/cp-2016-60, 2016
Manuscript under review for journal Clim. Past
Published: 20 June 2016
c
Author(s) 2016. CC-BY 3.0 License.
30
APPENDIX A. ESMARK’S METEOROLOGICAL TABLES IN 888
DEN NORSKE RIGSTIDENDE. 889
890
Esmark, J. 1818/19. Meteorologiske Iagttagelser i Christiania 1818, anstillede af 891
Prof. Esmark. Den Norske Rigstidende 1818, No. 7 (24 January); No. 10 (4 892
February); No. 14 (18 February); No. 18 (4 March); No. 23 (21 March), No. 893
28 (8 April), No. 32 (22 April); No. 37 (9 May); No. 40 (20 May), No. 45 (6 894
June), No. 49 (20 June), No. 54 (8 July); No. 59 (25 July); No. 63 (8 895
August); No. 67 (21 August); No. 71 (5 September); No. 83, (17 October); 896
No. 84 (21 October), No. 86 (28 October); No. 88 (4 November); No. 95 (28 897
November); No. 98 (9 December); No. 102 (23 December); No. 3 (8 January 898
1819). 899
Esmark, J. 1819/20. Meterologiske Iagttagelser i Christiania 1819, anstillede af 900
Prof. Esmark. Den Norske Rigstidende No. 6 (19 January); No. 11 (5 901
February); No. 16 (23 February); No. 19 (5 March); No. 24 (23 March); No. 902
26 (6 April); No. 33 (23 April); No. 36 (4 May); No. 41 (21 May); No. 48 903
(15 June); No. 49 (18 June); No. 54 (6 July); No. 62 (3 August); No. 65 (13 904
August); No. 67 (20 August); No. 78 (28 September); No. 79 (1 October) 905
No. 82 (12 October); No. 84 (19 October); No. 89 (5 November); No. 95 (26 906
November); No. 99 (10 December); No. 103 (24 December); No. 2 (7 907
January 1820). 908
Esmark, J. 1820/21. Meteorologiske Iagttagelser i Christiania 1820, anstillede af 909
Prof. Esmark. Den Norske Rigstidende, No. 7 (25 January); No. 11 (8 910
February), No. 14 (18 February); No. 18 (3 March); No. 24 (24 March) ; No. 911
28 (7 April); No. 32 (21 April); No. 37 (9 May); No. 41 (23 May); No. 47 912
(13 June); No. 50 (23 June); No. 54 (7 July); No. 58 (21 July); No. 63 (8 913
August); No. 68 (25 August); No. 72 (8 September); No. 77 (26 September); 914
No. 81 (10 October); No. 85 (24 October); No. 88 (3 November); No. 94 (24 915
November); No. 98 (8 December); No. 103 (26 December); No. 3 (9 January 916
1821). 917
Esmark, J. 1821/22. Meteorologiske Iagttagelser i Christiania 1821, anstillede af 918
Professor Esmark. Den Norske Rigstidende, No. 7 (23 January), står bare 919
snee,men ikke mengde, ; No. 11 (6 February); No. 16 (23 February); No. 21 920
(13 March); No. 23 (20 March); No. 29 (10 April); No. 33 (24 April), No. 38 921
Clim. Past Discuss., doi:10.5194/cp-2016-60, 2016
Manuscript under review for journal Clim. Past
Published: 20 June 2016
c
Author(s) 2016. CC-BY 3.0 License.
31
(11 May); No. 41 (22 May); No. 45 (5 June); No. 52 (29 June); No. 55 (10 922
July); No. 58 (20 July); No. 63 (6 August); No. 68 (24 August); No. 72 (7 923
September); No. 76 (21 September); No. 80 (5 October); No. 85 (22 924
October); No. 89 (5 November); No. 93 (19 November)(nytt moderne 925
plusstegn); No. 98 (7 December); No. 102 (21 December); No. 2 (7 January 926
1822). 927
Esmark, Jens 1822/23. Meteorologiske Iagttagelser i Christiania 1822, anstillede 928
ved Professor Esmark. Den Norske Rigstidende, No. 5 (18 January); No. 10 929
(4 February); No. 15 (22 February); No. 18 (4 March); No. 23 (22 March); 930
No. 28 (8 April); No. 32 (22 April); No. 36 (6 May); No. 42 (27 May); No. 931
45 (7 June) not nedbørmåling; No. 50 (24 June); No. 81 (11 October); No. 82 932
(14 October); No. 83 (18 October); No. 84 (21 October); No. 87 (1 933
November); No. 89 (8November); No. 90 (11 November); No. 92 (18 934
November); No. 94 (25 November); No. 96 (2 December); No. 98 (9 935
December); No. 102 (23 December); No. 2 (6 January 1823). 936
Esmark, J. 1823/24. Meteorologiske Iagttagelser i Christiania 1823, anstillede ved 937
Professor Esmark. Den Norske Rigstidende No. 7 (24 January); No. 11 (7 938
February) ; No. 15 (21 February); No. 20 (10 March); No. 24 (24 March); 939
No. 27 (4 April); No. 31 (18 April); No. 36 (5 May); No. 40 (19 May); No. 940
46 (9 June); No. 49 (20 June); No. 75 (19 September); No. 76 (22 941
September); No. 77 (26 September); No. 78 (29 September); No. 79 (3 942
October); No. 81 (10 October); No. 82 (13 October); No. 84 (20 October); 943
No. 88 (3 November); No. 93 (21 November); No. 98 (8 December); No. 102 944
(22 December); No. 2 (5 January 1824). 945
Esmark, J. 1824/25. Meteorologiske Iagttagelser i Christiania 1824, anstillede ved 946
Professor Esmark. Den Norske Rigstidende No. 6 (19 January); No. 11 (5 947
February); No. 15 (19 February); No. 20 (8 March); No. 24 (22 March); No. 948
29 (8 April); No. 33 (22 April); No. 37 (6 May); No. 42 (24 May); No. 45 (3 949
June); No. 50 (21 June); No. 54 (5 July); No. 59 (22 July); No. 64 (9 950
August); No. 68 (23 August); No. 74 (13 September); No. 77 (23 951
September); No. 80 (4 October); No. 86 (25 Oktober); No. 89 (4 November); 952
No. 96 (29 November); No. 98 (6 December); No. 103 (23 December); No. 2 953
(6 Januar 1825). 954
Clim. Past Discuss., doi:10.5194/cp-2016-60, 2016
Manuscript under review for journal Clim. Past
Published: 20 June 2016
c
Author(s) 2016. CC-BY 3.0 License.
32
Esmark, J. 1825/26. Meteorologiske Iagttagelser i Christiania 1825, anstillede ved 955
Professor Esmark. Den Norske Rigstidende No. 7 (24 January); No. 11 (7. 956
February), No. 15 (21 February); No. 18 (3. March); No. 24 (24 March); No. 957
29 (11 April); No. 33 (25 April); No. 36 (5 May); No. 40 (19 May); No. 45 958
(6 June); No. 49 (20 June); No. 53 (4 July); No. 70 (1 September); No. 71 (5 959
September); No. 73 (12 September); No. 74 (15. September); No. 76 (22 960
September); No. 79 (3 October), No. 85 (24 October); No. 89 (7 November); 961
No. 93 (21 November); No. 97 (5 December); No. 102 (22 December); No. 2 962
(5 January 1826). 963
Esmark, J. 1826/27. Meteorologiske Iagttagelser i Christiania 1826, anstillede ved 964
Professor Esmark. Den Norske Rigstidende No.8 (26 January); No. 12 (9 965
February); No. 17 (27 February); No. 19 (6 March); No.23 (20 March); No. 966
28 (6 April); No. 33 (24 April); No. 36 (4 May); No. 43 (29 May); No. 45 (5 967
June); No. 50 (22 June); No. 55 (10 July): No.58 (20 July); No. 62 (3 968
August); No. 67 (21 August); No. 72 (7 September); No. 77 (25 September); 969
No. 80 (5 Oktober); No. 84 (19 October); No. 88 (2 November); No. 93 (20 970
November); No. 97 (4 December); No. 102 (21 December); No. 2 (4 January 971
1827). 972
Esmark, J. 1827/28. Meteorologiske Iagttagelser i Christiania 1827, anstillede ved 973
Professor Esmark. Den Norske Rigstidende , No. 7 (22 January); No. 11 (5 974
February); No. 16 (22 February); No. 19 (5 March); No. 24 (22 March); No. 975
28 (5 April); No. 32 (19 April); No. 37 (7 May); No. 43 (28 May); No. 48 976
(14 June); No. 50 (21 June); No. 54 (5 July); No. 58 (19 July); No. 79 (1 977
October); No. 80 (4 October); No. 81 (8 October); No. 82 (11 October); No. 978
83 (15 October); No. 84 (18 October); No. 89 (5 November); No. 94 (22 979
November); No. 97 (3 December); 102 (20 December); No. 2 (7 January 980
1828) – also sums up last ten years, compares with Stockholm, the coldest 981
years have been 1819 and 1820, the mildest 1822 and 1826. 982
Esmark, J. 1828/29. Meteorologiske Iagttagelser i Christiania 1828, anstillede ved 983
Professor Esmark. Den Norske Rigstidende , No. 6 (21 January); No. 10 (4 984
February); No. 15 (21 February); No. 18 (3 March); No. 24 (24 March); No. 985
27 (3 April – mange solpletter); No. 32 (21 April); No. 36 (5 May); No. 40 986
(19 May); No. 45 (5 June); No. 49 (19 June); No. 53 (3 July); No. 59 (24 987
July); No. 63 (7 August); No. 78 (29 September); No. 79 (2 October); No. 81 988
Clim. Past Discuss., doi:10.5194/cp-2016-60, 2016
Manuscript under review for journal Clim. Past
Published: 20 June 2016
c
Author(s) 2016. CC-BY 3.0 License.
33
(9 October); No. 84 (20 October); No. 88 (3 November); No. 94 (24 989
November); No. 98 (8 December); No. 102 (22 December); No.2 (5 January 990
1829). 991
Esmark, J. 1829/30. Meteorologiske Iagttagelser i Christiania 1829, anstillede ved 992
Professor Esmark. Den Norske Rigstidende , No. 8 (26 January); No. 11 (5 993
February); No. 15 (19 February); No. 19 (5 March – den strengeste vinter på 994
mange år); No. 24 (23 March); No. 27 (2 April); No. 33 (23 April); No. 37 (7 995
May); No. 42 (25 May); No. 46 (8 June); No. 50 (22 June); No. 54 (6 July); 996
No. 78 (28 September); No. 79 (30 September); No. 80 (5 October); No. 81 997
(8 October); No. 85 (22 October); No. 87 (29 October); No. 89 (5 998
November); No. 90 (9 November); No. 94 (23 November); No. 99 (10 999
December); No. 103 (24 December); No. 2 (7 January 1830). 1000
Esmark, J. 1830/31. Meteorologiske Iagttagelser i Christiania 1830, anstillede ved 1001
Professor Esmark. Den Norske Rigstidende, No. 7 (25 January); No. 11 (8 1002
February); No. 14 (18 February); No. 18 (4 March); No. 22 (18 March); No. 1003
27 (5 April); No. 31 (19 April); No. 36 (6 May); No. 40 (19 May); No. 46 (9 1004
June); No. 50 (23 June); No. 53 (5 July); No. 57 (19 July); No. 63 (9 1005
August); No. 70 (1 September); No. 73 (13 September); No. 78 (29 1006
Septmerber); No. 81 (11 October); No. 84 (21 October); No. 91 (15 1007
November); No. 95 (29 November); 98 (9 December); No. 102 (23 1008
December); No. 3 (10 January 1831). 1009
Esmark, J. 1831/32. Meteorologiske Iagttagelser i Christiania 1831, anstillede ved 1010
Professor Esmark. Den Norske Rigstidende , No. 10 (3 February); No. 11 (7 1011
February); No. 17 (28 February); No. 20 (10 March); No. 25 (28 March); No. 1012
28 (7 April); No. 33 (25 April); No. 39 (12 May); No. 43 (22 May); No. 52 1013
(12 June); No. 57 (23 June); No. 63 (7 July); No. 70 (24 July); No. 75 (4 1014
August); No. 85 (28 August); No. 88 (4 September); No. 97 (25 September); 1015
No. 102 (10 October); No. 110 (3 November); No. 112 (10 November); No. 1016
118 (1 December); No. 119 (4 December); No. 1 (1 January 1832) ; No. 2 (5 1017
January 1832). 1018
Esmark, J. 1832/33. Meteorologiske Iagttagelser i Christiania 1832, anstillede ved 1019
Professor Esmark. Den Norske Rigstidende, No.10 (2 February); No. 11 (5 1020
February); No. 19 (4 March); No. 20 (8 March); No. 26 (26 March); No. 30 1021
(12 April); No. 33 (22 April); No. 37 (6 May); No. 43 (20 May); No. 52 (10 1022
Clim. Past Discuss., doi:10.5194/cp-2016-60, 2016
Manuscript under review for journal Clim. Past
Published: 20 June 2016
c
Author(s) 2016. CC-BY 3.0 License.
34
Juni); No. 57 (21 Juni); No. 63 (5 July); No. 70 (22 July); No. 78 (9 August); 1023
No. 86 (28 August – usedvanlig kold sommer); No. 92 (11 September); No. 1024
98 (25 September); No. 103 (7 October); No. 108 (25 October); No. 111 (4 1025
November); No. 117 (25 November); No. 122 (13 december); No. 127 (30 1026
December); No. 4 (13 Januery 1833). 1027
Esmark, J. 1833/34. Meteorologiske Iagttagelser i Christiania 1833, anstillede ved 1028
Professor Esmark. Den Norske Rigstidende, No.10 (3 February); No. 12 (10 1029
February); No. 18 (3 March); No. 24 (24 March); No. 25 (28 March); No. 30 1030
(14 April); No. 35 (2 May); No. 37 (9 May); No. 44 (26 May); No. 50 (9 1031
June); No. 58 (27 June); No. 63 (9 July); No. 77 (11 August); No. 80 (18 1032
August); No. 86 (1 September); No. 91 (12 September); No. 97 (26 1033
September); No. 103 (13 October); No. 105 (20 October); No. 110 (7 1034
November); No. 115 (24 November); No.120 (12 December); No. 123 (22 1035
December); No. 2 (5 January 1834). 1036
Esmark, J. 1834/35. Meteorologiske Iagttagelser i Christiania 1834, anstillede ved 1037
Professor Esmark. Den Norske Rigstidende ,No. 7 (23 Januery); No. 10 (2 1038
February); No. 16 (23 February); No. 18 (2 March); No. 24 (23 March); No. 1039
27 (3 April); No. 32 (20 April); No. 37 (4 May); No. 43 (18 May); No. 53 1040
(10 June); No. 60 (26 June); No. 68 (15 July)(regnet som falt på en 1041
kvadratfods flate utgjorde 4 rhinlandskae tommer eller 576 kubikktommer); 1042
No. 71 (22 July); No. 79 (10 August), No. 83 (19 August); No. 90 (7 1043
September); No. 96 (21 September); No. 102 (5 October); No. 107 (23 1044
October); No. 111 (6 November); No. 117 (27 November); No. 119 (4 1045
December); No. 126 (28 December); No. 2 (8 January 1835). 1046
Esmark, J. 1835/36. Meteorologiske Iagttagelser i Christiania 1835, anstillede ved 1047
Professor Esmark. Den Norske Rigstidende, No. 10 (1 February); No. 12 (8 1048
February); No.15 (19 February); No. 20 (8 March); No. 24 (22 March); No. 1049
28 (5 April); No. 34 (26 April); No. 40 (10 May); No. 50 (2 June); No. 54 1050
(11 June); No. 58 (21 June); No. 65 (7 July); No. 72 (23 July); No. 79 (9 1051
August); No. 88 (30 August); No. 91 (6 September); No. 99 (24 September); 1052
No. 105 (11 October); No. 107 (18 October); No. 112 (5 November); No. 1053
118 (26 November); No. 120 (3 December); No. 126 (24 December); No. 3 1054
(10 January 1836). 1055
Clim. Past Discuss., doi:10.5194/cp-2016-60, 2016
Manuscript under review for journal Clim. Past
Published: 20 June 2016
c
Author(s) 2016. CC-BY 3.0 License.
35
Esmark, J. 1836/37. Meteorologiske Iagttagelser i Christiania 1836, anstillede ved 1056
Professor Esmark. Den Norske Rigstidende, No. 7 (24 January); No. 15 (21 1057
February); No. 17 (28 February); No. 19 (6 March); No. 23 (20 March); No. 1058
27 (3 April); No. 32 (21 April); No. 38 (5 May); No. 45 (22 May); No. 50 (2 1059
June); No. 59 (23 June); No. 66 (10 July); No. 70 (19 July); No. 78 (7 1060
August); No. 85 (23 August?) ; No. 92 (8 September); No. 98 (22 1061
September); No. 105 (9 October); No. 111 (30 October); No. 112 (3 1062
November); No. 119 (27 November); No. 125 (18 December); No. 126 (22 1063
December); No. 3 (5 January 1837). 1064
Esmark, J. 1837/38. Meteorologiske Iagttagelser i Christiania 1837, anstillede ved 1065
Professor Esmark. Den Norske Rigstidende, No. 10 (22 January); No. 17 (7 1066
February); No. 22 (19 February); No. 22 (2 March); No. 34 (19 March); No 1067
41 (4 April); No. 48 (20 April); No. 53 (2 May); No. 61 (21 May); No. 67 (4 1068
June); No. 74 (20 June); No. 82 (9 July); No. 86 (18 July); No. 93 (3 1069
August); No. 100 (20 August); No. 106 (3 September); No. 113 (19 1070
September); No. 120 (5 October); No. 126 (19 October); No. 132 (2 1071
November); No. 139 (19 November); No. 145 (3 December); No. 152 (19 1072
December); No. 2 (4 January 1838). 1073
Esmark, J. 1838. Meteorologiske Iagttagelser i Christiania 1838, anstillede ved 1074
Professor Esmark. Den Norske Rigstidende, No. 10 (18 January); No. 19 (3 1075
February); No. 29 (20 February); No. 36 (4 March); No. 45 (20 March); No. 1076
53 (3 April); No. 62 (19 April); No. 70 (3 May); No. 79 (19 May); No. 87 (2 1077
June); No. 98 (19 June); No. 108 (4 Junly); No. 117 (19 July); No. 127 (2 1078
August); No. 137 (19 August); No. 148 (6 September); No. 156 (20 1079
September); No. 164 (4 October); No. 173 (20 October); No. 181 (3 1080
November); No. 190 (18 November); No. 199 (4 December); No. 207 (18 1081
December). 1082
Appendix"B"1083
MET Norway calculates monthly mean temperatures for manual stations by Føyn’s and 1084
Köppen’s formulas (Birkeland, 1936; Gjelten et al., 2014; Nordli et al., 2015), so we chose to 1085
use those formulas also for Esmark’s observations: The monthly mean temperature, T, may be 1086
calculated by Føyn’s formula and a modified Köppen’s formula, Table A1. 1087
Clim. Past Discuss., doi:10.5194/cp-2016-60, 2016
Manuscript under review for journal Clim. Past
Published: 20 June 2016
c
Author(s) 2016. CC-BY 3.0 License.
36
Table A1. Formulas for calculation of monthly mean temperature, T, where T08, T15 and T21, are 1088
monthly means at observation times 08, 15 and 21 UTC respectively, and Tn is monthly mean night 1089
temperature, kg and kf are constants. 1090
1091
Føyn’s formula
)( 15 ggg TTkTT −+=
2
2108 TT
Tg
+
=
Köppen’s formula
)TT(kTT nff −−=
2
2115 TT
Tf
+
=
1092
A “true” monthly mean temperature, T, may be calculated by the arithmetic mean of hourly 1093
observation according to definition, so for a station that have hourly observations the 1094
constants, kg and kf, are easily calculated by rearranging Føyn’s and Köppen’s formulas. For 1095
Esmark’s series from Øvre Vollgate the constants were unknown. It was assumed that the 1096
constants from Blindern could be used also for Øvre Vollgate. An indication of the robustness 1097
of this assumption was tested by comparison with a short series of hourly observations from 1098
the station 18815 Oslo – Bygdøy, 15 m a.s.l. The test procedure started with calculation of the 1099
constants for the Blindern series based on the period 2012.12-2015.09. These constants were 1100
then used for the calculation of mean monthly temperatures for Bygdøy for the same period, 1101
which were compared with the “true” monthly means, i.e. those calculated by the hourly 1102
observations. For Føyn’s formula the deviation from the true means varied from -0.06ᵒC in 1103
December to +0.18ᵒC in March that gave +0.05ᵒC for the whole year. Corresponding figures 1104
for Köppen’s formula were -0.06ᵒC in July, +0.16ᵒC in September and +0.01ᵒC for the whole 1105
year. These differences are so small that the lack of exact knowledge of the constants does 1106
add practically no uncertainty to the monthly temperatures. 1107
1108
Fig. 12. Differences in mean monthly temperature between Esmark’s observations 1109
at Øvre Vollgate and those at the Astronomical Observatory (Esmark minus 1110
Observatory) during the period 1837.04-1838.12. Esmark’s observations are 1111
presented both unadjusted and adjusted. For the observatory the temperatures 1112
are unadjusted. 1113
Clim. Past Discuss., doi:10.5194/cp-2016-60, 2016
Manuscript under review for journal Clim. Past
Published: 20 June 2016
c
Author(s) 2016. CC-BY 3.0 License.
37
1114 1115
Fig. 1. Map of Christiania (now Oslo) 1811 with the location of Esmark’s house in 1116
Øvre Vollgt. 7 marked with red star. 1117
1118
1119
1120
1121
1122
1123
Clim. Past Discuss., doi:10.5194/cp-2016-60, 2016
Manuscript under review for journal Clim. Past
Published: 20 June 2016
c
Author(s) 2016. CC-BY 3.0 License.
38
1124 1125
Fig. 2. Matriculation and survey 1830 of Esmark’s property No. 308, Øvre Voldgate 7, in 1126
Oslo Byarkiv (City archives). Garden to the left, house surrounding back yard to the right. 1127
1128
1129
1130
1131 1132
Fig. 3. Street view of Esmark’s house in Øvre Voldgate 7. Photograph from around 1900. The 1133
higher houses on both sides are late 19th century. Oslo Bymuseum, No. OB.F00897. 1134
1135
Clim. Past Discuss., doi:10.5194/cp-2016-60, 2016
Manuscript under review for journal Clim. Past
Published: 20 June 2016
c
Author(s) 2016. CC-BY 3.0 License.
39
1136 1137
Fig. 4. Esmark’s Christiania protocol for 1817. Now deposited at Riksarkivet 1138
(National archives), Oslo. S-1570. Det norske meteorologiske institutt. F/Fa. 1139
Materiale etter professorer. L0002. 1140
1141
1142 Fig. 5. The January page from Esmark’s meteorological observation protocol from 1143
1823, the year he discovered ice ages. Now deposited at Riksarkivet 1144
(National archives), Oslo. S-1570. Det norske meteorologiske institutt. F/Fa. 1145
Materiale etter professorer. L0002. 1146
Clim. Past Discuss., doi:10.5194/cp-2016-60, 2016
Manuscript under review for journal Clim. Past
Published: 20 June 2016
c
Author(s) 2016. CC-BY 3.0 License.
40
1147
1148 Fig. 6. The first published Christiania weather table, from Den norske Rigstidende, 1149
24 January 1818. 1150
1151
1152
1153
1154
1155
1156
1157
Fig. 7. The temperature 1158 difference (ᵒC) between
Esmark’s evening 1159 observation and the
morning observation the 1160 following day for the winter
season (Dec-Feb). 1161
Clim. Past Discuss., doi:10.5194/cp-2016-60, 2016
Manuscript under review for journal Clim. Past
Published: 20 June 2016
c
Author(s) 2016. CC-BY 3.0 License.
41
1162
1163
1164
1165
1166
1167
1168
1169
1170
1171
1172
1173
!"#$
%&'$
(")$
*+)$
(",$
!-#$
!-.$
*-/$
0&+$
123$
456$
7&2$
9:9$
9:9$
9:9$
89:;$
8<:;$
8=:9$
8<:?$
89:B$
9:9$
9:9$
9:9$
9:9$
1174
Fig. 8 Temperature differences (ᵒC) between the observations at Blindern at 15 UTC and at 21 UTC for the 1175
period 1993.01-2015.09. Also the difference between the midday and evening observations of Esmark is shown 1176
for the period 1816.01-1838.12. (The adjustments of the evening observations, Table 5, are added to the data for 1177
the period 1816.01-1821.12 before the calculation of the differences. In the table below the figure are shown the 1178
adjustments of Esmark’s midday observation. 1179
1180
Clim. Past Discuss., doi:10.5194/cp-2016-60, 2016
Manuscript under review for journal Clim. Past
Published: 20 June 2016
c
Author(s) 2016. CC-BY 3.0 License.
42
1181
1182
1183 1184
Fig. 9. Adjustments added to Esmark’s series for each season during his period of 1185
observation, 1816-1838. 1186
1187
Clim. Past Discuss., doi:10.5194/cp-2016-60, 2016
Manuscript under review for journal Clim. Past
Published: 20 June 2016
c
Author(s) 2016. CC-BY 3.0 License.
43
1188
Fig. 10. Annual and seasonal means of Esmark’s temperature series (symbols), and 1189
Gaussian filter (curves) with standard deviation 3 in the Gaussian distribution C&:/:$45)D.E$1190
&3$".:F$=9<AG, corresponsing roughly to a 10 year regtangular filter. 1191
1192
Clim. Past Discuss., doi:10.5194/cp-2016-60, 2016
Manuscript under review for journal Clim. Past
Published: 20 June 2016
c
Author(s) 2016. CC-BY 3.0 License.
44
1193
1194
1195 1196
Fig. 11. Difference between Esmark’s observations at Øvre Vollgate and Hansteen’s 1197
observations at Pilestredet (Esmark minus Hansteen) during the period 1822.11-1827.02 at 1198
08, 15 and 21 UTC. The monthly means are calculated by Føyn’s formula (see Appendix B). 1199
1200
Clim. Past Discuss., doi:10.5194/cp-2016-60, 2016
Manuscript under review for journal Clim. Past
Published: 20 June 2016
c
Author(s) 2016. CC-BY 3.0 License.
45
1201
1202 1203
Fig. 12. Differences in mean monthly temperature between Esmark’s observations at Øvre 1204
Vollgate and those at the Astronomical Observatory (Esmark minus Observatory) during the 1205
period 1837.04-1838.12. Esmark’s observations are presented both unadjusted and adjusted. 1206
For the observatory the temperatures are unadjusted. 1207
1208
1209
1210
1211
1212
1213
1214
1215
1216
Clim. Past Discuss., doi:10.5194/cp-2016-60, 2016
Manuscript under review for journal Clim. Past
Published: 20 June 2016
c
Author(s) 2016. CC-BY 3.0 License.
