Trends in climate variables and their combined effects on irrigation water demand in the dry season

Abstract

Agriculture is a key economic sector in Bangladesh and one of the main users of fresh watcr resources. At prcsent, climate change is thought to be putting extra stress on agricultural productivity and hence food security because climate change may lead to changes in irrigation water requirement. Most of the studies rclated to the etfeets of climate change on irrigation water requirement have been done based only on changes of temperature and/or rainfall. Other climatic variables, such as solar radiation, humidity and wind speed, which may also ,,[[ect the irrigation watcr requirement, were not taken into consideration in those studies. To find out the impact of climate change on agricultural water requiremcnt, the combined effects of all the variables need to be considered. This study is based on secondary ditta nnd information. Datn of four meteorological station~, namely Dhaka, Jessore, Bogra and Chandpur, on dnily observations of air temperature, air humidity, wind speed, solar radiation and minfall arc used. The stations are selected such thnt one station is selected from eueh of the North-Central, South-West, North-West imd South-East hydrological planning regions of Bangladesh. IO-daytime scale for data analysis has been used as it closcly resembles agricultural and water resources planning activitie> in Bangladesh. The Penm;ll1- Monteith method is used for reference crop evapotranspiration (ETo)computation and both parametric and non-parametric methods are used for testing thc statistical significance of trends in different hydro-climatic variables. The results of the analysis reveal that maximum temperature has i~ereasing trcnds of 0.2~C, 0.2GCand 0.1~C per decade foJ' Dhaka, Jcssore and Chnndpur, rcspectively, during the dry season (November-May), In contrast, Bogra has a deercasing trcnd of 0.2oCper decade. Minimum temperatLirehas increasing trends of a.30C, a.2oC, 0.2GC and O_l~Cper decade at Dhaka, Jessore, Bogra and Chandpur, reopectively. The rdative humidity has inerea"ing trends at four Mationsin aU 10-day periods of the dry season. The average increasing trends are 0.9%, 2.5%, 3,0% and 1.4% per dccade at Dhaka, Jessore, Bogra and Chandpur, respectively_ The other impoltant elimatic vminble is the sunshine hour which has decreasing trends ill aliI O-dayperiods at the four stations. The average decreasing trends at Dhaka, Jcssore, Bogra and Chandpur stutions are 0.7, 0.5, 0.5 and 0.7 hours per decade, respectively, Wind speed has both decreasing and increasing trends depending on the stations, The average decreasing trends ill wind speed at Dhaka and Chandpur arc 15.8 and 40,8 km/day per decade, respectively. The average increasing trends in wind speed are 5.5 kmlday and 1.1 km/day per decade at lessore and 13ogra,respectively. From the lrend analysis of elimatic variables at four stations, it is found thm sunshine hour has decreasing trend and relative humidity, maximum temperature and minimum temperature have increasing trends. The combined effeet of the trends of above climatic variables on ETo and net irrigation requirement (NIR) of Dora rice, which is the staple crap in Bungladesh, are lhen evuluated. The ET0 is found to have decreasing trends at all stations. The decreasing trends are 0.2, 0.2, 0.1 and 0.3 mm/day per decade (10 years) at Dhaka, Jessorc, Bogra and Chandpur, respectively, during the dry season. The average decrease oC ETo from 1961 to 2007 is found to be about 23%, 25%, 8% and 33% at Dhaka, Jessore, Bogra and Chandpnr, respectively. The net irrigation requirements (NIR) for Bora rice were estimated from rice evapotranspiration and rainfall data. Decreasing trends in NIR of 0.1,0.1,0.1 and 0.4 mmfday per decade at Dhaka, lessore, Bogra and Chandpur stations, respectively, are found, Seasonal mean NIR of 1991-2007 is found to be lower than thai of 1961-1975. It appears from this study that the combined effect oCtlle climatic variables oftlle four ~lations is the decreasing trends of EToand NIR. Besides, the rainfall has increasing lrcnds at the four stations which also contribute to the decreasing trend in NIR. Though the temperatures have increasing trends which should increase ET0and NIR, the effects of other variables arc more dominating than that of tcmperatures. The findings of this study convey a clear message to the policy and decision makers lind water managers that the irrigation sector would not bc adversely affected due 10 climate change. These findings are in contrary to the gcncral belief und reported modeling results which state that due to climate changc water requirement for irrigation would increase,