Red River CMIP5 bias correction and downscaling

Python and R code for bias correction and empirical-statistical downscaling of CMIP5 projections for the Red River basin in Vietnam. Bias correction and downscaling of CMIP5 projections was conducted in support of:

Quinn JD, Reed PM, Giuliani M, Castelletti A, Oyler JW, Nicholas RE. 2018. Exploring How Changing Monsoonal Dynamics and Human Pressures Challenge Multi-Reservoir Management for Flood Protection, Hydropower Production and Agricultural Water Supply. Water resources research. DOI: 10.1029/2018WR022743.

Bias Correction Methods

We use a variation of the modified equidistant quantile matching (EDCDFm) algorithm (Li et al. 2010) of (Pierce et al. 2015) to bias correct CMIP5 projections for the Red River, Vietnam. Details of the Red River variation of EDCDFm are provided here.

Downscaling Methods

We use a variation of the constructed analogs method described by (Pierce et al. 2014) to downscale CMIP5 projections for the Red River, Vietnam. Our modified method is termed Constructed Analogs with Single Anomaly Analog (CASAA). Details of the Red River CASAA method are provided here.