Improving PERSIANN-CCS Rainfall Estimates using Probability Matching Method (PMM)

Satellite remotely sensed data has wide application in monitoring and evaluating hydrological and meteorological events such as precipitation, floods, droughts, and climate extremes. Low Earth Orbit Satellites (LEO) and Geostationary Operational Environmental Satellites (GOES) are the two most commonly used satellites for these purposes. Despite the high temporal and spatial resolution of GOES data, they are an indirect method for estimating the rainfall rate since they measure the brightness temperature of clouds from Infrared (IR) channels or the cloud albedo from visible (VIS) sensors. Passive Microwave (PMW) sensors can detect information about cloud characteristics and cloud formation for precipitation estimates. Hence, regardless of the sparse spatial and temporal resolution they can produce more reliable precipitation data. This study looks at the combined precipitation products from LEO-PMW and GEO-IR estimations from the global point of view. Precipitation Estimation from Remotely Sensed Information using Cloud Classification System (PERSIANN-CCS), is an IR-based precipitation algorithm that uses image segmentation and feature extraction to classify cloud images into a number of clusters based on the neural network method.The probability matching method (PMM) implemented calculates the bias correction on CCS data using PMW  estimation. The model was calibrated using climatological data form (2008 to 2011) and validated over the entire year of 2012. The results show that the improvement is more significant over the high latitude regions and less significant over the low latitude regions. The bias has significantly decreased in the global scale by using this approach.