Vol.3, No.1B, 16-19 (2014) Journal of Agricultural Chemistry and Environment
http://dx.doi.org/10.4236/jacen.2014.31B004
Modeling crop land soil moisture and impacts of
supplimental irrigaiton in a rainfed region of
Bangladesh
Pramod K. Pandey*, Sagor Biswas
Department of Population Health and Reproduction, University of Californian, Davis, USA;
*Corresponding Author:
Received October 213
ABSTRACT
A robust water balance model has been tested
for predicting soil moisture levels and supple-
mental irrigation requirement of a rainfed region
of Bangladesh. The predictions were used for
improving the understanding of the impacts of
rainwater harvesting on rainfed agriculture. The
climate data (i.e., rainfall, temperature, evapora-
tion, and evapotranspiration) were used as in-
puts for predicting the variations in soil mois-
ture. Soil moisture levels under rainfed and sup-
plementary irrigation conditions were compared.
Results showed that rainwater harvesting i.e.,
rain water storage tanks during rainy seasons
can be potentially useful for storing rainwater,
which can be utilized for enhancing crop land
soil moisture during dry seasons for enhancing
crop yield. The study presented here will be use-
ful for improving and disseminating rainwater
harvesting approaches for enhancing water
availability in rainfed regions.
KEYWORDS
Rainwater Harvesting; Rainfed Crop Land;
Supplemental Irrigation; Crop Yiel d
1. INTRODUCTION
It is required to increase the agricultural water availa-
bility in rainfed regions to enhance the global food pro-
duction. Approximately more than 80% of the global
crop land is rainfed, which produces more than 70% of
global food productions currently [1-3]. For improving
food production further, additional water resources capa-
ble of providing the irrigation to crop lands is required
[4]. One option is increasing the facilities/structures for
rainwater harvesting in the crop land itself [5,6]. In many
rainfed regions, for instance, in Bangladesh, more than
76% of rainfall oc curs in rainy season (May to October);
however, a major portion of it losses as runoff. Due to
insufficient water storages, farmers often face irrigation
water shortages during dry seasons. Providing the facili-
ties capable of storing the rain water during rainy season
can potentially facilitate water availability for irrigation.
Previous studies have shown that harvested rainwater in
on-farm reservoirs during rainy season can enhance crop
yield considerably [7,8]. Here we have exploited a water
balance model [8] for calculating soil moisture and crop
yield under rainwater harvesting facilities and without
rainwater harvesting (i.e., rainfed) for improving the un-
derstanding of rainfed agriculture and rainwater harvest-
ing approaches.
2. METHODS
2.1. Study Area
The study area is show n in F i gure 1 . Jessore, a distr ict
situated in the southwestern part of Bangladesh (BD),
receives about 1741 mm of annual rainfall. Nearly 76%
of annual rainfall occurs from May to October. Out of
that about 28% of the total annual rainfall occurs in the
month of July. Temperature varies from 10˚C to 36˚C.
Relative humidity varies from 72% to 86%, and wind
speed varies from 0.76 to 4.6 m/s. The rainfall and tem-
perature variations of the study area are shown in Fig-
ures 2 and 3. Average monthly evapotranspiration varia-
tion is shown in Figure 4.
2.2. Model
The model used in the study has been described else-
where [8]. The model has two components: 1) water bal-
ance simulation for crop land; and 2) water balance si-
mulation for water storage tanks. Water storage tanks re-
ceive water from upland catchment area of 5 ha (as ru-
noff), and direct precipitation on tank’s surface. The
stored water in the tanks was applied as supplemental
irrigation (when needed) to the crop land for enhancing

P. K. Pandey, S. Biswas / Journal of Agricultural Chemistry and Environment 3 (2014) 16-19
17
Figure 1. Study area (Jessore District, Bangladesh).
Figure 2. Precipitation (Jessore District, Bangladesh).
Figure 3. Temperature (Jessore District, Bangladesh).
soil moisture. In simulation, we used crop land area of 1
ha, reservoir area of 15% of the catchment area (i.e., 0.8
ha).
Figure 4. Potential evapotranspiration (Jessore, BD).
The model uses curve number for estimating the ru-
noff from catchment to water storage tanks. Daily see-
page, evaporation, and spill from the tank were simulated,
and the simulation methods are described previously [5,
8]. Crop water requirement were predicted using readily
available soil moisture and non-readily available soil
moisture [8]. The crop coefficient of bean crop was used
for simulating the crop water requirement at various crop
growth stages as described previously [8].
The model requires multiple input parameters, which
are described in two previous studies [5,8]. Readers are
encouraged to preview the published studies for under-
standing the model’s details. In this study, we used rainfall,
temperature, and evapotranspiration data from the Jes-
sore District of Bangladesh. The average monthly rainfall
and min/max temperatures were obtained from Bangla-
desh Agricultural Research Council (BARC) [9]. Using
the monthly data, we estimated daily data using polyno-
mial equations (fitted on monthly data). Due to unavaila-
bility of evapotranspiration data of the Jessore District,
we used neighboring climate stations for estimating the
evaporation and evapotranspiration for the study area. The
data of the neighboring stations (i.e. West Bengal, India)
were obtained from two sources: 1) Indian Meteorologi-
cal Department (IMD) [10], and 2 ) Hydrology and Water
Resources Information System for India [11]. The climate
data (i.e., temperature and rainfall) of the location in India
were compared with the Jessore District, and the data
were comparable. For example, the annual monthly rain-
fall data had similarity of 78% and annual monthly tem-
perature had the similarity of 87%. After combining the
deviation of rainfall and temperature, we anticipate that
there was a possibility of 12.7% deviation in climates
between the study area and neighboring climate station.
3. CALCULATIO NS AND INPUT DATA
To estimate the daily precipitation from the average
monthly data shown in Figure 2, we did perform two
separate interpolations: 1) rising limb of precipitation;
and 2) falling limb of precipitation, which yielded daily
0
200
400
600
800
1000
1 3 5 7 911
Precipitaiton (mm)
Months
2005 2006 2007 2008
2009 2010 Average
15
19
23
27
31
35
1357911
Temperature (d eg C)
Months
2005 2006 2007 2008
2009 2010 Average
y = 0.2965x3-8.1653x2+ 59.066x + 7.1564
R² = 0.8126
0
40
80
120
160
200
0246810 12 14
Potential evapotranspiratoin monthly (mm)
Months

P. K. Pandey, S. Biswas / Journal of Agricultural Chemi s try and Environment 3 (2014) 16-19
18
precipitation data (Figure 5). Two separate interpolations
were needed because single interpolation was not able to
capture the peak rainfall, which occurred in the month of
July. The rising and falling limbs of precipitation are
shown in Figure 5, and interpolated precipitation values
(daily) are also shown in the figure. Similarly, daily eva-
potranspiration was estimated using the average monthly
shown in Figure 4. The daily evapotranspiration is
shown in Figure 6.
The model used in this study requires daily input data
(precipitation, temperature, evaporation, and evapotrans-
piration). Evaporation was estimated from evapotranspi-
ration. Previous studies reported that evaporation values
vary approximately 120% - 130% of evapotranspiration.
In this study, we used daily evaporation values as 130%
of the dail y ev a potranspira t ion values.
4. RESULTS AND DISCUSSION
The average annual interpolated daily precipitation
was slightly less than the observed data. The observed
average annual precipitation for the study area was about
1741 mm, while interpolation yielded average annual
precipitation of 1478 mm i.e., 81% of the observed val-
ues. The average annual evapotranspiration was approx-
imately 1189 mm, while interpolation yielded average
annual evapotranspiration of 1147 mm. Figure 7 shows
the soil moisture variations in rainfed and irrigated con-
ditions. In addition, daily precipitation and supplemental
irrigation is also shown in the figure. The simulation is
shown for starting from Julian Day 1 to Julian Day 150.
As shown in the figure, soil moisture was considerably
elevated when supplemental irrigation was applied (sup -
plemental irrigation is shown as vertical red bars in Fig-
ure 7). Soil moisture in the rainfed and irrigated condi-
tions were estimated for two seasons: (Season 1: Julian
Day 20 - 119; and Season 2: Juli a n Da y 16 5 - 264). Com-
pared to the first season, in the second season i.e., be-
yond Julian Day 165, soil moisture in rainfed and irrigat-
ed conditions were comparable because of excess rainfall.
The available water storages in ponds were not utilized
as supplemental irrigation because the soil moisture was
suffi cient without supplem ental irrigation ( data not shown).
At the end of cropping season, the soil moisture con-
tent in irrigated condition was almost three times greater
than the rainfed soil moisture. Although actual evapo-
transpiration (ET) was almost two times greater in irri-
gated condition compared to rainfed condition, the actual
yield increased about three times in irrigated condition
when compared to the rainfed condition. Addition of 128
mm of supplementary irrigation decreased the green wa-
ter use by 45% and increased the total water use by 55%
compared to rainfed condition. Subsequently the overall
water use efficiency showed a nearly 55% increase in
irrigated condition. As shown in the Table 1, water re-
Figure 5. Interpolated daily precipitation.
Figure 6. Daily evapotranspiration.
Figure 7. Soil moisture and supplemental irrigation.
charge values (R) of uncultivated land in irrigated condi-
tion was 85% of the recharge value of rainfed fed condi-
tion. ET value of uncultivated land (i.e., catchment) in
irrigated condition was 86% of the rainfed condition. In
cultivated land, R and ET values in irrigated condition
were higher than the rainfed condition. For example, R
y = 4E-06x
3
-0.0006x
2
+ 0.0377x -0.1228
R² = 0.9839y = -3E-06x
3
+ 0.0028x
2
-0.9952x + 126.7
R² = 0.9403
0
5
10
15
20
25
30
0
5
10
15
20
25
30
050100 150 200 250300 350 400
Interpolated precipitaiton (mm)
Precipitation (mm)
Julian days
Rising limb of rainfallFalling limb of rainfall
Interpolated precipitaitonPoly. (Rising limb of rainfall)
Poly. (Falling limb of rainfall)
0
2
4
6
050100 150 200 250300 350 400
Potential Evapotranspiration (mm)
Julian Days
0
4
8
12
16
20
24
20
60
100
140
180
020406080100 120 140 160
Precipitation (mm)
Soil moisture (mm)
Julian Days
Supplimental irrigationSoil moisture with irrigation
Soil moisture with rainfedPrecipitation

P. K. Pandey, S. Biswas / Journal of Agricultural Chemistry and Environment 3 (2014) 16-19
19
Table 1. Water balance parameters of catchment area, culti-
vated land, and crop yields.
Conditions
Parameters
AMe
(mm) ETa (mm) Ya/Ym Ya (kg/ha) TS (mm) GW (mm)
Irr. 90.2 268.7 1.0 6000 128 73.53
Rain. 33.2 130.5 0.36 2141 0.0 130.48
Parameters
OWUE
(kg/m3)
UA CA
R (m3/yr) ET (m3/yr) R (m3/yr) ET (m3/yr)
Irr. 2.98 28101 21588 5907 7842
Rain. 1.64 33060 25397 5498 6298
**Note: Irr. = irrigated; rain. = rainfed; AMe = availab le moistur e at the end
of croppi ng seas on 1; Ya = actual cr op yiel d; Y m = maximum crop yield; TS
= total supplemental irrigation applied; GW = green water use; OWUE =
overall water use efficiency; R = recharge; ET = Evapotranspiration; UA =
Uncultivated land area (catchment area of 5 ha); CA = Catchment area
(cultivated land area of 1 ha).
value in rainfed condition was 93% of the irrigated con-
dition, and ET value in rainfed condition was approx-
imately 80% of the irrigated condition. In summary, the
results of the study showed that rainwater harvesting
approach can be an effective alternative for enhancing
agricultural water availability in the rainfed regions.
5. CONCLUSION
A water balance model was used to estimate the im-
pacts of rainwater harvesting appr o ach on enhancing
rainfed crop land soil moistures and crop yield for a
southwestern district of Bangladesh. The model esti-
mated rainwater storages in water storage tank (designed
in the farm land). The model uses algorithms to estimate
the water requirement of the crop land as well as water
availability in the tanks. This decision making allows
model to estimate the supplemental irrigation require-
ment in the crop land as well as supplemental irrigation
availability in the tanks. There sults showed th at the r ain-
water harvesting approach presented here increased crop
yield considerably in the studied rainfed region of the
Bangladesh. The model requires four major parameters:
precipitation, temperature, evaporation, and evapotrans-
piration. To run the model, daily input data are required.
In this study, daily data were estimated from the availa-
ble monthly data and used to feed the model. We antic-
ipate that the availability of daily observed data will im-
prove the model predictions significantly, therefore, fur-
ther studies utilizing the daily observed data for predict-
ing supplemental irriga tion, soil moisture, and crop yield
will be necessary. We suggest future studies utilizing the
climate data of multiple locations (rainfed) to enhance
the model as well as model predictions.
ACKNOWLEDGEMENTS
We than k to Dr. Pieter van der Zaag, Professor, UNESCO-IHE Insti-
tute for Water Education, Delft, Netherlands, Water Resources Section,
Delft University of Technology, Delft, Netherlands for his help in mod-
el development, which is already published elsewhere [8].
REFERENCES
[1] Human Development Report (2010) The real wealth of
nations: pathways to human development. United Nations
Development Program (UNDP).
http://hdr.undp.org/en/media/HDR_2010_EN_Complete_
reprint.pdf
[2] Rockström, J., Falkenmark, M., Karlberg, L., Hoff, H.,
Rost, S. and Gerten , D. (2009) Future water availability
for global food production: The potential of green water
for increasing resilience to global change. Water Resourc-
es Research, 45, Article ID: W00A12.
[3] Rockström, J., Barron, J. and Fox, P. (2003) Water prod-
uctivity in rain-fed agriculture: Challenges and opportun-
ities for smallholder farmers in drought-prone tropical
agro-ecosystems. In: Kijne, J.W., Barker, R. and Molden,
D., Eds., Water Productivity in Agriculture: Limits and
Opportunities for Improvements, CABI, Publ., Oxon,
145-162.
[4] World Bank (2010) Water resource management. Wash-
ington, DC.
http://web.worldbank.org/WBSITE/EXTERNAL/TOPIC
S/EXTWAT/0,,contentMDK:21630583~menuPK:460244
5~pagePK:148956~piPK:216618~theSitePK:4602123,00.
html
[5] Pandey, P.K., Soupir, M.L., Singh, V.P., Panda, S.N. and
Pandey, V. (2011) Modelling rainwater storage in distri-
buted reservoir systems in humid subtropical and tropical
savannah regions. Water Resources Management, 25,
3091-3111. http://dx.doi.org/10.1007/s11269-011-9847-5
[6] Panigrahi, B., Panda, S.N. and Mull, R. (2001) Simula-
tion of water harvesting potential in rainfed ricelands us-
ing water balance model. Agricultural Systems, 69, 165-
182. http://dx.doi.org/10.1016/S0308-521X(01)00013-0
[7] Pandey, P.K., Panda, S.N. and Panigrahi, B. (2006) Sizing
on-farm reservoirs for crop-fish integration in rainfed
farming systems in Eastern India. Biosystems Engineer-
ing, 93, 475-489.
ttp://dx.doi.org/10.1016/j.biosystemseng.2006.01.009
[8] Pandey, P.K., der Zaag, P., Soupir, M.L. and Singh, V.P.
(2013) A new model to simulate hydro-economic poten-
tial of rainwater harvesting and supplemental irrigation
rainfed agriculture. Water Resources Management, 27,
3145-3164. http://dx.doi.org/10.1007/s11269-013-0340-1
[9] (2013) Bangladesh Agricultural Research Council, Min-
istry of Agriculture, Government of Peoples Republic of
Bangladesh. http://www.barc.gov.bd/inf_sys_data.php
[10] (2013) Indian Meteorological Department (IMD).
http://www.imd.gov.in/
[11] (2013) Hydrology and Water Resources Information Sys-
tems for India. http://www.nih.ernet.in/rbis/rbis.htm