Erosion and Accretion Index for Kuwaiti Coast

Abstract

Assessment of long term erosion and accretion rate of the coastal area is essential for selection of different types of coastal structures. For example, it is not advisable to develop housing infrastructure on the coast, which is historically eroding. Similarly it is not preferable to select an accreting site for the construction of a port in order to avoid annual maintenance dredging after construction of the port. Kuwait is a coastal country and hence it is essential to understand the long term coastal morphological changes of the coast and identify the historically accreting or eroding or stable coast over a period of many years, say 10 to 15 years. The coastline evolution of Kuwait is investigated using Landsat remote sensing images (30 m resolution) from 1989 to 2003, IRS-P5 images (2.5 m resolution) obtained for the year 2006-07 and 2003 aerial photos (0.38 m resolution). A total of 130 grids, each at a spacing of 1 nautical mile is used. The extent of accretion/erosion for 17 years (from 1989 to 2006) for the mainland coastline of Kuwait is assessed. The average shoreline change/year is calculated. Erosion and accretion index for each location is assigned based on the average annual erosion and accretion. From this study, it is found that annual erosion of more than 25 m does not exist in Kuwait. 1.54% of the coast has experienced annual erosion of 10 to 25 m and 0.77% of the coast has experienced 0 to 10 m annual erosion. 3.85% of the coast is stable. 43.85% of the coast has annual accretion of 0 to 10 m, 11.54% has annual accretion of 10 to 25 m and 38.46% has annual accretion of greater than 25 m. The complete details of this study are presented in this paper. The results will be useful for integrated management of Kuwaiti coast.

Full text

Int. J. Environ. Res., 7(3):679-684,Summer 2013
ISSN: 1735-6865
Received 9 Oct. 2012; Revised 20 Dec. 2012; Accepted 12 March 2013
*Corresponding author E-mail:
nsubram@kisr.edu.kw
679
Erosion and Accretion Index for Kuwaiti Coast
Neelamani, S.
*
and Uddin , S.
Environmental and Urban Development Division, Kuwait Institute for Scientific Research, P.O.
Box: 24885, 13109 Safat, Kuwait
ABSTRACT: Assessment of long term erosion and accretion rate of the coastal area is essential for selection
of different types of coastal structures. For example, it is not advisable to develop housing infrastructure on
the coast, which is historically eroding. Similarly it is not preferable to select an accreting site for the
construction of a port in order to avoid annual maintenance dredging after construction of the port. Kuwait is
a coastal country and hence it is essential to understand the long term coastal morphological changes of the
coast and identify the historically accreting or eroding or stable coast over a period of many years, say 10 to
15 years. The coastline evolution of Kuwait is investigated using Landsat remote sensing images (30 m
resolution) from 1989 to 2003, IRS-P5 images (2.5 m resolution) obtained for the year 2006-07 and 2003 aerial
photos (0.38 m resolution). A total of 130 grids, each at a spacing of 1 nautical mile is used. The extent of
accretion/erosion for 17 years (from 1989 to 2006) for the mainland coastline of Kuwait is assessed. The
average shoreline change/year is calculated. Erosion and accretion index for each location is assigned based on
the average annual erosion and accretion. From this study, it is found that annual erosion of more than 25 m
does not exist in Kuwait. 1.54% of the coast has experienced annual erosion of 10 to 25 m and 0.77% of the
coast has experienced 0 to 10 m annual erosion. 3.85% of the coast is stable. 43.85% of the coast has annual
accretion of 0 to 10 m, 11.54% has annual accretion of 10 to 25 m and 38.46% has annual accretion of greater
than 25 m. The complete details of this study are presented in this paper. The results will be useful for
integrated management of Kuwaiti coast.
Keywords: Erosion, Accretion, Stable coast, Kuwaiti shoreline, Coastline evolution, Remote sensing,
Natural and manmade coastal changes
INTRODUCTION
Coastal erosion or accretion is considered as
hazards for infrastructure developments in the coastal
area. For example, if a coastal area is eroding
systematically over a period of many years, then the
sea will occupy the coastal infrastructure and destroy
the foundations of such structures. It is a hazard for
such type of coastal structures. On the other hand, if
sand/silt is accreting systematically, then it will a stroke
of luck and provide a wide beach and good for
recreation. In certain areas, sand/silt accumulation is
not wanted. For example, if sand is accumulating
around seawater intake structure sites, it is a hazard.
So one cannot say that erosion is hazard and accretion
is a piece of good fortune. It depends upon the type of
coastal project activity. Kuwait is a coastal country
(Fig.1) and has a total coastal length of 496 km. Most of
the population lives closer to coast. The latitude and
longitude at its southern tip is N 28
o
32' 8.16'’ and E 48
o
26' 00". The northern tip of the country is near N 29
o
59'
and E 48
o
00' 48’’ as sh own in Fig.1.
The government is also planning for many more
infrastructures in the available coastal area. Hence a
general understanding of whether a stretch of Kuwaiti
coast is historically eroding or accreting will be useful
for selecting a suitable site for different types of
coastal project like Marina, Port and Harbors, Public
Beaches, Marine outfalls from power and desalination
plants, Culverts and natural drains, Sea Water In takes,
Piers, Moles, Beach Nourishments etc. The coastline
evolution of Kuwait is studied using Landsat remote
sensing images (30 m resolution) from 1989 to 2003,
IRS-P5 images (2.5 m resolution) obtained for the year
2006-07 and 2003 aerial photos (0.38 m resolution).
The extent of accretion/erosion during the year 1989-
2006 for the whole coastline of Kuwait is brought out.
Most of the man made coastal developments is focused
from the Southern Kuwaiti coastal boundary to
Shuwaikh Port and hence the coastal morphological
changes are found to be mainly due to the responses
of these developments. The developments of sand
spits in many parts of southern coastal areas were

680
Neelamani,S. and Uddin, S.
Fig.1. Latitude and Longitude details of Kuwait
observed. Accretions in most of the southern Kuwaiti
coasts are found to be due to the influence of slipways
which are acting as groin fields. A detailed
comprehensive geo-environmental data base is
prepared (Neelamani et al., 2009). The scientific details
used for the analysis of remote sensing images are
also described in detail in Neelamani et al., 2009.
A proper understanding of the extent of erosion
and accretion and the spatial variation is essential for
planning future projects. It is hence useful to create an
index, which will reflect whether a coastal area is stable
or has low/medium/high erosion rate or low/medium/
high accretion rate, using index values for better
understanding. This paper gives such information.
MATERIALS & METHODS
The change deductions of the coastline are
obtained through superposition of remote sensing
images belong to 1989 an d 2003 as well as 2006 -07. The
tidal corrections were applied for the correct change
deduction of the coast. An index value of 0 is assigned
for stable coastline. Low/medium/high coastal erosion
is given i ndex value of -1, -2 and -3 respectivel y. Index
value of -1 is assigned if the annual rate of erosion is 0
to 1 0m, -2 for ann ual er osion in between 10 t o 25 m an d
-3 for annual erosion rate of greater than 25 m. Similarly
low/medium/high accretion rates are assigned with
index value of 1, 2 and 3 respectively. Index value of 1
is assigned if the annual rate of accretion is 0 to 10m, 2
for annual accretion in between 10 to 25 m and 3 for
ann ual accret ion rat e of greater than 25 m. The foll owing
section describes the details of the hazard assessment
index purely based of the order of average rate of annual
erosion and accretion for the Kuwaiti coastal area. Th e
hazard index values are provided for a total of 130
locations along the Kuwaiti coast. This will give a
firsthand impression of the order of annual erosion
and accretion occurred at these sites during 1989 to
2006-07.
RESULTS & DISCUSSION
The hazard assessment is carried out purely based
of the order of average rate of annual erosion and
accretion in Kuwaiti coastal area. The information is
gathered from the change detection of coastline
obtained by superposing the coastlines (after tide
correction) of landsat image of 1989 and IRS P5 images
of 2006-07. The maximum value of erosion/accretion
on each gri d of 1’ x 1’ is ex tracted. The average shorelin e
change per year is calculated. An erosion/accretion
index is formed based on different ranges of annual
rate of erosion or accretion as explained in Table 1.
Annual shoreline
change (m) Meaning Index
Value
Accretion greater
than 25 m High
accretion rate 3
Accretion in the
range of 10 to 25 m Medium
accretion rate 2
Accretion in the
range of 0 to 10 m Low accretion
rate 1
No change in
coastline Sta ble c oast 0
Erosion in the range
of 0 to 10 m Low erosion
rate -1
Erosion in the range
of 10 to 25 m Medium
erosion rate -2
Erosion greater than
25 m High erosion
rate -3
Table 1. Erosi on and Accretion Index Based on
Annual Rate of Change in Kuwaiti Shoreline
Table 2 to 7 gives the index values for each 1’ x 1’ grid
for different coastal segments in Kuwait.
Table 2 shows the erosion and accretion index
based on shoreline change in Kuwait during 1989 to
2006 from Nuwaisib to Ras Al-Julayah Covering Al-
Khiran and Az-Zour. Since the southern most Kuwaiti
coast (28
o
32’ 8.16" N to 28
o
38’ N) has high accretion
rate during 1989-2006, it carries index value of 3. One
location has index value of -2, indicating medium
erosion rate during 1989-2006. Most of the locations
in this coastal stretch carry index value of 1, which
indicates that this coastal stretch from 28
o
42’ N to 28
o
51’ N has dominated by low accretion rate during 1989-
2006.

Int. J. Environ. Res., 7(3):679-684,Summer 2013
681
Table 3 shows the erosion and accretion index
based on shoreline change in Kuwait during 1989 to
2006 from Ras Al-Julayah to Al-Fintas covering
Shuaiba, Ahmadi, Fahaheel and Fintas coastal areas.
Latitude Longitude Tot al shoreline change
during 1989-2006 (m) Average shoreline
change/year ( m) Erosion/
Accretion index
28
o
32' 8.16'' 48
o
26' 00” 2170 127.6 3
28
o
33' 48
o
25' 12'' 2046 120.4 3
28
o
34' 48
o
24' 24'' 1612 94.8 3
28
o
35' 48
o
23' 55'' 3410 200.59 3
28
o
36' 48
o
23' 36'' 496 29.18 3
28
o
37' 48
o
23' 12'' 1426 83.88 3
28
o
38' 48
o
23' 40'' 496 29.18 3
28
o
39' 48
o
23' 33'' -310 -18.24 -2
28
o
40' 48
o
23' 24'' 372 21.88 2
28
o
41' 48
o
23'00” 0 0 0
28
o
42' 48
o
22' 48'' 124 7.29 1
28
o
43' 48
o
22' 54'' 124 7.29 1
28
o
44' 48
o
23' 26'' 1240 72.9 3
28
o
45' 48
o
20' 10'' 124 7.29 1
28
o
46' 48
o
18' 35'' 62 3.65 1
28
o
48' 48
o
17'00” 124 7.29 1
28
o
49' 48
o
16' 48'' 62 3.65 1
28
o
50' 48
o
16' 54'' 124 7.29 1
28
o
51' 48
o
16' 48'' 124 7.29 1
Table 2. Erosion and Accretion Index Based on Shoreline Change in Kuwait During 1989 to 2006 from
Nuwaisib to Ras Al-Julayah Covering Al-Khiran and Az-Zour
Table 3. Erosion and Accretion Index Based on Shoreline Change in Kuwait During 1989 to 2006 from Ras Al-
Julayah to Al-Fintas Covering Shuaiba, Ahmadi, Fahaheel and Fintas
Latitude Longitude Total shoreline change
during 1989-2006 (m) Average shoreline
change/year (m) Erosion/
Accretion index
28
o
52' 48
o
16' 54'' 186 10.94 2
28
o
53' 48
o
16' 00'' 124 7.29 1
28
o
54' 48
o
14' 00'' 62 3.65 1
28
o
55' 48
o
12' 36'' 62 3.65 1
28
o
56' 48
o
12' 30'' 62 3.65 1
28
o
57' 48
o
11' 48'' 62 3.65 1
28
o
58' 48
o
10' 42'' 62 3.65 1
28
o
59' 48
o
10' 18'' 62 3.65 1
29
o
00' 48
o
09' 54'' 62 3.65 1
29
o
01' 48
o
09' 36'' 62 3.65 1
29
o
02' 48
o
09' 30'' 0 0 0
29
o
03' 48
o
09' 18'' 0 0 0
29
o
04' 48
o
09' 00'' 62 3.65 1
29
o
05' 48
o
08' 24'' 62 3.65 1
29
o
06' 48
o
08' 15'' 93 5.47 1
29
o
07' 48
o
08' 09'' 62 3.65 1
29
o
08' 48
o
08' 00'' 124 7.29 1
29
o
09' 48
o
07' 42'' 62 3.65 1
29
o
10' 48
o
07' 30'' 62 3.65 1
29
o
11' 48
o
07' 12'' 124 7.29 1
Most of the locations carry index value of 1, which
indicates that this coastal stretch has dominated by
low accretion rate durin g 1989-2006. Th e low accretion
is due to slipways built by the population living in this

682
Erosion and Accretion Index for Kuwaiti Coast
Table 4. Erosion and Accretion Index Based on Shoreline Change in Kuwait During 1989 to 2006 from
Al-Fintas to Ras Al-Ardh
Latitude Longitude Total shoreline change
during 198 9-2006 (m) Average shoreline
change/year (m) Erosion/
Accretion index
29
o
12' 48
o
06' 54'' 186 10.94 2
29
o
13' 48
o
06' 15'' 62 3.65 1
29
o
14' 48
o
06' 06'' 62 3.65 1
29
o
15' 48
o
05' 48'' 31 1.82 1
29
o
16' 48
o
05' 30'' 62 3.65 1
29
o
17' 48
o
05' 24'' 62 3.65 1
29
o
18' 48
o
05' 24'' 62 3.65 1
29
o
19' 48
o
05' 24'' 62 3.65 1
29
o
20' 48
o
05' 36'' 62 3.65 1
29
o
21' 48
o
06' 12'' 62 3.65 1
coastal area, for handling their pleasure boats. Two
locations (29
o
02’ N to 29
o
03’ N) car ry index value of 0
indicating stable coastline. These two locations has
Shuaiba power station and Mina As-Shuaiba Port, build
using concrete structures and the coastal ar e protected
from possible erosion using ripraps.
Table 4 shows the erosion and accretion index
based on shoreline change in Kuwait during 1989 to
2006 from Al-Fintas to Ras Al-Ardh. Most of the
locations carry index value of 1, which indicates that
this coastal stretch has dominated by low accretion
rate during 1989-2006. The low accretion is due to the
presence of many manmade coastal structures.
Table 5 shows the erosion and accretion index
based on shoreline change in Kuwait during 1989 to
2006 in South Kuwait Bay-From Ras Al-Ardh to Jahra
Beach area. Ras Al-Ardh to Shuwaikh port area shows
low accretion. Most of this coastal stretch is man-
made. Artificial beach formations using sand
nourishments were carried out during the last few
decades and has reflected in this index. The outer
side of Shuwaikh port (47
o
55’ E) shows erosion index
Table 5. Erosion and Accretion Index Based on Shoreline Change in Kuwait During 1989 to 2006 in South
Kuwait B ay-From Ras Al-Ardh to Jahr a Beach Area
Latitude Longitude Total shoreline change
during 1989-2006 (m) Average shoreline
change/year (m) Erosion/Accretion
index
29
o
21' 12'' 48
o
06' 0 0 0
29
o
20' 48'' 48
o
05' 62 3.65 1
29
o
20' 36'' 48
o
04' 62 3.65 1
29
o
20' 36'' 48
o
03' 62 3.65 1
29
o
21' 00'' 48
o
02' 124 7.30 1
29
o
22' 06'' 48
o
01' 62 3.65 1
29
o
23' 30'' 48
o
00' 62 3.65 1
29
o
23' 30'' 47
o
59' 0 0 0
29
o
22' 56'' 47
o
58' 62 3.65 1
29
o
21' 30'' 47
o
57' 62 3.65 1
29
o
21' 10'' 47
o
56' -124 -7.3 -1
29
o
21' 27'' 47
o
55' -496 -29.18 -2
29
o
20' 00'' 47
o
54' 930 54.71 3
29
o
19' 36'' 47
o
53' 496 29.18 3
29
o
19' 12'' 47
o
52' 682 40.12 3
29
o
19' 12'' 47
o
51' 930 54.71 3
29
o
22' 36'' 47
o
50' 1240 72.94 3
29
o
22' 30'' 47
o
49' 62 3.65 1
29
o
23' 06'' 47
o
48' 62 3.65 1
29
o
22' 36'' 47
o
47' 62 3.65 1
29
o
22' 06'' 47
o
46' 310 18.24 2
29
o
22' 06'' 47
o
45' 868 51.1 3
29
o
21' 36'' 47
o
44' 930 54.7 3
29
o
21' 12'' 47
o
43' 744 43.76 3
29
o
21' 36'' 47
o
42' 558 32.82 3

683
Int. J. Environ. Res., 7(3):679-684,Summer 2013
Table 6. Erosion and Accretion Index Based on Shoreline Change in Kuwait During 1989 to 2006 in Kuwait
Bay- Northern Part
Latitude Longitude Total shoreline change
during 1989-2006 (m) Average shoreline
change/year (m) Er osio n/
Accretion index
29
o
22' 48'' 47
o
42' 372 21.8 8 2
29
o
23' 48'' 47
o
43' 124 7.29 1
29
o
24' 10'' 47
o
44' 372 21.8 8 2
29
o
24' 12'' 47
o
45' 620 36.4 7 3
29
o
25' 24'' 47
o
46' 558 32.8 2 3
29
o
26' 24'' 47
o
47' 558 32.8 2 3
29
o
27' 48'' 47
o
48' 992 58.3 5 3
29
o
28' 48'' 47
o
49' 105 4 62.0 3
29
o
30' 00'' 47
o
50' 161 2 94.82 3
29
o
30' 48'' 47
o
51' 192 2 113.06 3
29
o
31' 24'' 47
o
52' 148 8 87.53 3
29
o
31' 48'' 47
o
53' 111 6 65.64 3
29
o
32' 24'' 47
o
54' 111 6 65.64 3
29
o
32' 54'' 47
o
55' 111 6 65.64 3
29
o
34' 00'' 47
o
56' 111 6 65.64 3
29
o
36' 24'' 47
o
57' 136 4 80.24 3
29
o
37' 00'' 47
o
58' 167 4 98.47 3
29
o
37' 24'' 47
o
59' 142 6 83.88 3
29
o
37' 36'' 48
o
00' 930 54.7 1 3
29
o
37' 24'' 48
o
01' 868 51.0 6 3
29
o
37' 00'' 48
o
02' 744 43.7 6 3
29
o
35' 48'' 48
o
03' 868 51.0 6 3
29
o
35' 36'' 48
o
04' 682 40.1 2 3
29
o
35' 24'' 48
o
05' 930 54.7 1 3
29
o
33' 12'' 48
o
07' 806 47.4 1 3
29
o
32' 30'' 48
o
08' 682 40.1 1 3
29
o
32' 00'' 48
o
09' 142 6 83.88 3
29
o
31' 40'' 48
o
10' 105 4 62.0 3
29
o
32' 36'' 48
o
11' 434 25.5 3 3
of -2, which most probably due to dredging activities
around this location. Most of the locations in the
western part of Shuwaikh port is tidal flat an d indicates
high accretion with index of 3. This can be linked with
long term sediment deposit as a reaction of forced
deccication of Mesopotamian marshland in Iraq.
Table 6 shows the erosion and accretion index
based on shoreline change in Kuwait during 1989 to
2006 in the Northern Kuwait Bay stretch. Most of the
locations in the this part of Kuwaiti coast is tidal flat
and indicates high accretion with index of 3. Again,
this can be linked only with long term sediment
deposition of the sediments reached into Kuwait Bay
due to forced empting of the water and sediments of
15,000 km
2
of the Mesopotamian marshland in Iraq
during the study period. Further detailed study is
required.
Table 7 shows the erosion and accretion index
based on shoreline change in Kuwait during 1989 to
2006 in the mainland side of Khor As-Subiya. This
channel is very dynamic and tide induced currents are
significant. The accretion index value varies from 1 to
3.The accretion can be contributed to the sediment
supply from the Mesopotamian marshland and supply
of sediments through Shatt Al-Arab and Shatt Al-Basra.
In general, the areas around river mouth is expected to
grow due to the supply of sediments and the variation
in the accumulation is controlled by the hydrodynamic
forces due to the tidal variations.
Table 2 to 7 is useful as a general guideline to identify
areas of erosion and accretion and their intensity in
the Kuwaiti coastal area.
More information on Kuwaiti coastline evolution can
be obtained from Neelamani et al. (2007) and Neelamani
and Saif uddin (2009a, 2009b, 2 010 and 2011).

684
Neelamani,S. and Uddin, S.
Table 7. Erosion and Accretion Index Based on Shoreline Change in Kuwait During 1989 to 2006 in Khor As-
Subiya-Mainland Side
Latitude Longitude Total shoreline change
during 1989-2006 (m) Average shoreline
change/year (m) Erosi on/Ac cretion
index
29
o
33' 48
o
11' 12’’ 124 7.29 1
29
o
34' 48
o
10' 42’’ 744 43.76 3
29
o
35' 48
o
09' 54’’ 186 10.94 2
29
o
36' 48
o
09' 24’’ 124 7.29 1
29
o
37' 48
o
09' 00’’ 186 10.94 2
29
o
38' 48
o
08' 24’’ 62 3.65 1
29
o
39' 48
o
08' 00’’ 62 3.65 1
29
o
40' 48
o
07' 24’’ 62 3.65 1
29
o
41' 48
o
07' 00’’ 62 3.65 1
29
o
42' 48
o
06' 24’’ 62 3.65 1
29
o
43' 48
o
06' 12’’ 310 18.24 2
29
o
44' 48
o
04' 48’’ 372 21.88 2
29
o
45' 48
o
03' 00’’ 434 25.53 3
29
o
46' 48
o
02' 30’’ 372 21.88 2
29
o
47' 48
o
02' 12’’ 186 10.94 2
29
o
48' 48
o
01' 48’’ 124 7.29 1
29
o
49' 48
o
01' 12’’ 124 7.29 1
29
o
50' 48
o
01' 00’’ 124 7.29 1
29
o
51' 48
o
00' 48’’ 124 7.29 1
29
o
52' 48
o
00' 48’’ 248 14.59 2
29
o
53' 48
o
01' 00’’ 310 18.24 2
29
o
54' 48
o
01' 12’’ 496 29.18 3
29
o
55' 48
o
00' 36’’ 1240 72.94 3
29
o
56' 48
o
00' 24’’ 1984 116.70 3
29
o
57' 48
o
00' 24’’ 620 36.47 3
29
o
58' 48
o
00' 24’’ 620 36.47 3
29
o
59' 48
o
00' 48’’ 310 18.24 2
CONCLUSION
Any country with coastline/shoreline should
know the general trend of shoreline changes for a long
period of time. Erosion and accretion are common
coastal phenomena. Both nature induced force and
man-made effect contribute for the shoreline changes.
The shoreline change for Kuwaiti coastal area is
studied using remote sensin g images from 1989 to 2006-
07. Most of the Kuwaiti coastal stretches show trend
of sediment accretion. Only few locations show
erosions during 1989-2006. The value of the erosion
and accretion index given in this paper may be of use
for identifying suitable coastal sites for different types
of future coastal projects. Further study is needed for
fine tuning of the present results.
ACKNOWLEDGEMENTS
The authors would like to acknowledge with thanks
Kuwait Foundation for the Advancement of Science
for funding this project work. Our appreciation for the
management of Kuwait Institute for Scientific Research
for providing the needed R&D facility for this research
project work.
REFERENCES
Neelamani, S., Saif ud din, Rakha, K., Zhao, Y., Al-Nassar,
W., Al-Banaa, K., Al-Salem, K., Al-Othman, A. an d Alragum,
A. (2007). “Coastline Evolution of Kuwait Using Remote
Sensing Techniques, EC022C. Final Report, KISR 9051.
Neelamani, S. and Saif ud din. (2009a). Investigation on
Coastline Evolution – A case study for Kuwaiti Coast.
International Conference in Ocean Engineering, Feb. 1-5,
IIT Madras, India, pp.721-734.
Neelamani, S. and Saif ud din. (2009b). Kuwaiti coastline
evolution during 1989 to 2003. Second Conference &
Exhibition on Environm
ent & Sustainability, 16-19,
Kuwait, pp.787-803.
Neelamani, S. an d Saif ud din. (2010). E volution of Coast in
the Vicinity of a Seawater Intake Structure in Kuwait.
International Journal of Imaging, 3, 3-22.
Neelamani, S., Saif uddin and Anbazagan, S. (2011 ). Kuwait
coastline Evolution during 1989-2007. Book chapter in
Geoinformatics in Applied Geomorphology, Chapter 5, CRC
Press, Taylor & Francis, USA, pp. 87-104.