Overcoming Economic Impact of Rainfall Variability in Mahailuppallama by Using Gamma Distribution

Abstract

Introduction : Climate change is a major factor that significantly affects the world economy with extreme changes in climatic variables such as precipitation, temperature, winds, relative humidity etc. Agriculture is an economic endeavor that depends on both climatic and weather situations on which its extremes direct to failure in agronomic activities switching deficit in the economy. Parallel circumstance exists in Sri Lanka, where the primary income source is agriculture.

With the high geographic and climatic variability in Sri Lanka, different kinds of crops are cultivated in different regions, harmonizing their unique characteristics of topography and climate. The mean annual rainfall in Sri Lanka varies from under 900mm to over 5000mm from the driest part to the wettest part. The 3 climatic zones that have been classified according to the mean annual rainfall are further divided into 46 Agro-ecological regions having different amount of rainfall.

Crops, cultivars and implementation of agronomic practices are planned as per the forecast climatic variables all over the country. In addition to that, irrigation planning is executed mainly considering rainfall where irrigated farming is performed. Thus, rainfall variability is crucial in agricultural activity, water management, food security and energy production. According to the annual reports of Central Bank of Sri Lanka from 2010 to 2019, paddy production in yala season of 2012, 2014, 2017, 2018 and 2019 has declined due to the drought conditions existing in those years whereas it has increased in 2010, 2011, 2013, 2015, 2016 and 2018 due to the favorable weather conditions and proper irrigation plannning. It generates fluctuations in the economy since mainly drought conditions prevail in the Dry Zone, where paddy production is concentrated in Sri Lanka. Therefore, realization of rainfall distribution is vital in order to accomplish proper policy planning, decision making and risk management.

Hence, location specific range and likelihood of rainfall is essential in achieving those strategies. Therefore, modeling of rainfall variability with probability distribution is a useful tool. The information regarding rainfall accumulation in time and space for an area and the foundation for fitting and testing distribution models is given by historical rainfall data (Husak et al., 2007). Along with that the gamma distribution has been recommended as the best fitted distribution in order to describe the annual, monthly or seasonal rainfall (Aksoy, 1999; Sen and Eljadid, 1999; Husak et al., 2007; Sivajothi and Karthikeyan, 2016).