SPATIOTEMPORAL VARIABILITY ANALYSIS OF RAINFALL IN BANGLADESH

Abstract

Bangladesh is a densely-populated, mainly riverine and low-lying country located in South Asia. The economy is agrarian and approximately 80% of its total population are involved in a varied sort of agricultural activities. The mean annual rainfall in Bangladesh is 2300 mm. Two large river systems, i.e. Ganga and Brahmaputra drain-off their water to the Bay of Bengal after traversing through the country. Due to this very fact, it is one of the most flood-affected countries in the world. Bangladesh is also characterized as one of the climate change susceptible countries in the world. It has been expected that rainfall and temperature will be intensified due to climate change which might affect the water resources sector, agricultural activities and public health etc. Information on the temporal and spatial distribution and patterns of rainfall is very important for a variety of uses in hydrology and water resources planning and management. APHRODITE gridded rainfall data of 0.25 x 0.25 degree resolution was analyzed to study spatiotemporal trends of rainfall and rainy days in Bangladesh during 1951-2007. Analysis of annual rainfall trends of 184 rainfall grids, covering entire Bangladesh was carried out using Mann–Kendall test and Sen’s slope test. Trend analysis revealed an increasing trend in the annual data (182 grids) as well as in the seasonal rainfall data for all seasons. Pettitt–Mann–Whitney test, used for detecting a change in a trend, highlighted the most probable year of rainfall trend change as 1982. A comparative analysis between observed data and APHRODITE data showed that the trends of annual as well as seasonal rainfall series for both the dataset matched in most cases. Most of the grids prior to 1982, depicted a decreasing trend in annual data, monsoon (156 grids) and post-monsoon (178 grids) data. But, an increasing trend was found in the partial data series of winter (183 grids) and pre-monsoon seasons (148 grids). After the trend change year i.e. 1982, an increasing trend was found in annual data (99 grids), and in monsoon (114 grids) and post monsoon (142 grids) data. A decreasing trend was observed in pre-monsoon (126 grids) and winter rainfall (184 grids) data. The trend analysis of rainy days have shown an increasing trend in annual, pre-monsoon and winter season data during 1951-2007. But, monsoon and post-monsoon seasons have revealed decreasing trend for the same duration. Similar analysis for both the partial series (prior and after 1982) revealed decreasing trend in iv annual, monsoon and post-monsoon rainy days. Pre-monsoon and winter rainfall showed an increasing trend during 1951-1982, followed by a decreasing trend during 1983-2007. The spatial distribution of the trends revealed the fact that the annual and monsoon rainfall have increased with a decrease in rainy days in lower parts of Bangladesh. Pre-monsoon rainfall has increased in the drought prone northwest part of Bangladesh and decreased in other parts. Winter rainfall has decreased all over the country. The spatial analysis of the partial series after the trend change indicated that the lower parts of Bangladesh are more vulnerable to the rainfall changes in comparison to other parts of the country. Besides, analysis of extreme rainfall event indicates that with the increasing rainfall, the extreme rainfall events are also increasing in the lower parts of Bangladesh which indicates that the lower parts of Bangladesh is more vulnerable to flooding. Correlation analysis of El Niño Southern Oscillation shows strong teleconnections with rainfall in Bangladesh. It was found that Bangladesh is prone to flooding during the La Niña and moderate El Niño years whereas the country has never been faced by any disastrous flood/cyclone during strong El Niño years. In the strong El Niño years a significant deficit in rainfall was also detected. Combined effect of rainfall, rainy days and extreme event trends indicates towards more intense rainfall events in coming future intensifying flood problems in Bangladesh. Decreasing trends in rainfall and rainy days in pre-monsoon and winter rainfall data indicates towards longer dry spells leading to droughts affecting agriculture practices in Bangladesh. It is expected that the increase in flooding problem as well as in dry spells will severely affect the economy and food production in Bangladesh.