Applications of palmer method in analyzing drought in northwest region of Bangladesh

Abstract

In the present study, an attempt has been made for applying the Palmcr mcthod in drought analysis in the Northwest (NW) region of Bangladesh. Palmer Drought Severity Index CPOSI) actually uses a supply and dcmand model for the amount of moisture in the soil. The method defined drought period as an interval of timc during which the actual moisture supply at a givcn place falls short of the climatically expected or appropriate moisturc supply. In the study the PDSI was calculated on monthly basis as per original formulation by Palmer, though a shorter timescale is more appropriate for irrigation scheduling in Bangladcsh. The drought analysis by the method showed that all of the stations of NW region experienced some degrees of drought conditions. The method quantified the average occurrence of drought in the region in 34 percent of the study time from 1972 to 2002. The available literature shows that thc average occurrence of drought is once in 2.5 years. The method identified the most common drought occurrences in years like 1972.73, 1975.76, 1976.77, 1978.79, 1979.80, 1982-83, 1992.93, 1994.95, 1999 and 2001.2002 in selected fifteen stations in the region. Thus the method could identitY the historical drought occurrences of Bangladesh like the year 1975, 1979, 1981.82, 1984, 1989, 1992.93 and 1994.95. The occurrences of extreme drought conditions were observed only in 3.6 percent of total time. In the study, thc Palmer method has identified the month of October rather than November or December as thc beginning of highest percentage of drought spells in the region. The method could identify Kharif droughts during the period from June/July to October. The method has identified the month of April rather than Mayor June as the ending of highest percentage of drought spells in the region. The suitability of the application of the Palmer method as a drought analyzing tool in NW region of Bangladesh has been assessed in the study by comparing the results of this method with those obtained from two other methods: i) the Standard Precipitation Index (SPl) method and ii) the Herbst method. The comparison was made in terms of several drought features such as the onset and termination of drought, duration and severity. The SPI is a standardized transform of the probability of the observed

precipitation. The method used half-monthly observed rainfall data to analyze drought on the monthly time scale. The Herbst method quantified droughts on a half-monthly time scale using different drought parameters such as mean half-monthly rainfall, weighting factors, carryover, mean half monthly deficit etc. The method has no drought severity scale rather provides relative measures of droughts. The average occurrence of drought was in 22.3 percent and 22 percent of study time in the SPI and the Herbst methods respectively. These two methods have also identified the same drought years in NW region as in the Palmer method in the present study. The highest percentage of beginning of drought spells occur in the months of March and November respectively in the SPI and Herbst methods while the highest percentage of ending of drought spell occurs in the month of April in the SPI method and in MaylJune in the Herbst method. While applying Palmer method in drought analysis of NW region of Bangladesh, some limitations were noticed in addition to those already identified by many researchers. Despite these limitations, thc performance of the method in comparison with the SPI and Herbst methods seems realistic in NW region of Bangladesh in tenns of drought occurrences and severity. But a more detailed study is recommended for quantitative assessment of the influence of soil moisture on drought occurrences in Bangladesh. The Palmer method can be used extensively for monitoring drought conditions and for making operational water management decisions in Bangladesh after some modifications to the method.