Future Rainfall Projection Using Modified Local Intensity Scaling (M-LOCI) Method.

Abstract Downscaling techniques are required because General Circulation Models (GCM) cannot be applied directly to climate change projections for climate change impact assessments. However, there are no definite guidelines for choosing a specific statistical downscaling technique to be applied to a certain application. The results of four separate RCPs (Representative Concentration Pathways), each driven by a different model, are scaled in this chapter to evaluate two statistical downscaling methods. Daily rainfall time series for a future scenario are generated for Dumdum station for the period 2021-50 to 2051-80 under climate change projections. Since the fundamental difference between downscaling methodologies relates to how severe events are produced, attention is paid to alterations of extreme events. However, 8 of the 10 GCMs investigated show that future precipitation occurrences will increase. The uncertainties are partially caused by the variability of the statistical downscaling methods and the projection. This chapter highlights the importance in annual average of different GCMs statistical downscaling methods and comparison between different models and different RCPs of same model for use in different hydrological methods.