Abstract
South Asian monsoon is a phenomena that plays out during June-September every year, due to the northward shift of the ITCZ which causes heavy rainfall over many countries of South Asia, including India. These rains are directly related to the lives and economic well-being of over a billion people. Indian monsoon is highly heterogeneous, due to the vast physiographic variations across the country. There is considerable interest among scientists and other stake-holders about possible future changes to Indian monsoon due to worldwide climate change. Simulations of future climate by global climate models under various scenarios can provide important clues for this. However, simulations of Indian monsoon in the historical period by global climate models under the CMIP5 family were found to be inaccurate in several aspects. Simulations by the new global climate models from the CMIP6 family are now available, and scientists are evaluating their ability to simulate Indian monsoon. In this work, we focus on one particular aspect of simulations by these models: the spatial distribution over daily rainfall over the Indian landmass during monsoon. We use a Machine Learning based probabilistic graphical model that can identify frequent spatial patterns of rainfall after creating a binary representation of rainfall. This model also helps us to identify spatial clusters, i.e., homogeneous regions within the Indian landmass with similar temporal characteristics of rainfall. We identify such frequent spatial patterns and spatial clusters from observed monsoon rainfall data, and also from simulations of monsoon rainfall by different CMIP6 models during the period 2000–2014. We evaluate the models by comparing the patterns and clusters identified from their simulations with those identified from observed data. We find that some of the CMIP6 models can simulate the spatial distribution of monsoon rainfall to a reasonable degree, but there are various limitations—most models underestimate extreme rainfall events and are unable to reproduce the regions of the landmass that are homogeneous with respect to rainfall.
Sign-in/Register to access full text options 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a callDisclaimer: All third-party content on this website/platform is and will remain the property of their respective owners and is provided on "as is" basis without any warranties, express or implied. Use of third-party content does not indicate any affiliation, sponsorship with or endorsement by them. Any references to third-party content is to identify the corresponding services and shall be considered fair use under The CopyrightLaw.
