Bayesian approach to Spatio-temporally Consistent Simulation of Daily Monsoon Rainfall over India

Simulation of rainfall over a region for long time-sequences can be very useful for planning and policy-making, especially in India where the economy is heavily reliant on monsoon rainfall. However, such simulations should be able to preserve known spatial and temporal characteristics of rainfall over India. General Circulation Models (GCMs) are unable to do so, and various rainfall generators designed by hydrologists using stochastic processes like Gaussian Processes are also difficult to apply over the highly diverse landscape of India. In this paper, we explore a series of Bayesian models based on conditional distributions of latent variables that describe weather conditions at specific locations and over the whole country. During parameter estimation from observed data, we use spatio-temporal smoothing using Markov Random Field so that the parameters learnt are spatially and temporally coherent. Also, we use a nonparametric spatial clustering based on Chinese Restaurant Process to identify homogeneous regions, which are utilized by some of the proposed models to improve spatial correlations of the simulated rainfall. The models are able to simulate daily rainfall across India for years, and can also utilize contextual information for conditional simulation. We use two datasets of different spatial resolutions over India, and focus on the period 2000--2015. We consider metrics to study the spatio-temporal properties of the simulations by the models, and compare them with the observed data to evaluate the strengths and weaknesses of the models.

[1]  C. W. Richardson.  Stochastic simulation of daily precipitation, temperature, and solar radiation , 1981 .

[2]  V. Isham,et al.  A point process model for rainfall: further developments , 1988, Proceedings of the Royal Society of London. A. Mathematical and Physical Sciences.

[3]  C. Glasbey,et al.  Rainfall Modelling Using a Latent Gaussian Variable , 1997 .

[4]  D. Wilks.  Multisite generalization of a daily stochastic precipitation generation model , 1998 .

[5]  Sulochana Gadgil,et al.  The Indian monsoon and its variability , 2003 .

[6]  M. Rivington,et al.  Assessment of the ClimGen stochastic weather generator at Cameroon sites , 2007 .

[7]  D. Watkins,et al.  Stochastic rainfall modeling in West Africa: Parsimonious approaches for domestic rainwater harvesting assessment , 2008 .

[8]  A. Pandey,et al.  Stochastic Modelling of Rainfall in Humid Region of North East India , 2008 .

[9]  Soumya Ghosh,et al.  Spatial distance dependent Chinese restaurant processes for image segmentation , 2011, NIPS.

[10]  R. Katz,et al.  Daily spatiotemporal precipitation simulation using latent and transformed Gaussian processes , 2012 .

[11]  V. V. Srinivas.  Regionalization of Precipitation in India – A Review , 2013 .

[12]  Yong-Sik Cho,et al.  A Poisson Cluster Stochastic Rainfall Generator That Accounts for the Interannual Variability of Rainfall Statistics: Validation at Various Geographic Locations across the United States , 2014, J. Appl. Math..

[13]  Bangwon Ko,et al.  Stochastic simulation of daily precipitation: A copula approach , 2014 .

[14]  Anastassia Baxevani,et al.  A spatiotemporal precipitation generator based on a censored latent Gaussian field , 2015 .

[15]  C. Jayasankar,et al.  Robust signals of future projections of Indian summer monsoon rainfall by IPCC AR5 climate models: Role of seasonal cycle and interannual variability , 2015 .

[16]  Balaji Rajagopalan,et al.  Are we unnecessarily constraining the agility of complex process‐based models? , 2015 .

[17]  Pierre Ailliot,et al.  Stochastic weather generators: an overview of weather type models , 2015 .

[18]  S. Karmakar,et al.  Do dynamic regional models add value to the global model projections of Indian monsoon? , 2017, Climate Dynamics.

[19]  B. Rajagopalan,et al.  A conditional stochastic weather generator for seasonal to multi-decadal simulations , 2018 .