Differentiation of Interpolation Techniques for Estimating the Spatial Distribution of Average Rainfall in Bangladesh

Authors: Mizanur Rahman*,
Topics: Geographic Information Science and Systems, Climatology and Meteorology
Keywords: GIS, Rainfall, Interpolation, Bangladesh
Session Type: Poster
Day: 4/4/2019
Start / End Time: 8:00 AM / 9:40 AM
Room: Lincoln 2, Marriott, Exhibition Level
Presentation File: Download

It is projected that heat-trapping gases in the atmosphere will change global precipitation patterns to become more extreme making dry seasons more dry and wet seasons wetter. In Bangladesh, the dry and wet seasons are getting longer. This affects roughly 80 percent of the total population in Bangladesh, which is directly or indirectly engaged in agricultural activities. In this study, four interpolation techniques (Empirical Bayesian Kriging, Global Polynomial Interpolation, Inverse Distance Weighted, and Radial Basis Function) were compared for estimating the spatial distribution of rainfall in Bangladesh. Rainfall values recorded from 34 weather stations were averaged from 2003-2013 to help remove any anomalies. 12 stations were randomly chosen and set aside for independent validation, and the remaining 22 stations were used to calibrate each model. Cross-validation using the Root Mean Square Error (RMSE) was used to determine the optimal parameters for four interpolation techniques. Empirical Bayesian Kriging achieved the best results where the Mean Bias Error (MBE) is -0.61 and RMSE is 3.32. Besides, spatial distribution, the temporal distribution of rainfall also important for regional climate change impact studies. Therefore, the analysis of the temporal distribution of rainfall is desired in future studies.

Abstract Information

This abstract is already part of a session. View the session here.

To access contact information login