UC Irvine

Earth observing satellites provide a method to measure precipitation from space with good spatial and temporal coverage. These estimates have a high degree of uncertainty associated with them. Understanding and quantifying the uncertainty of the satellite estimates can be very beneficial when using these precipitation products in hydrological applications given that these uncertainties will propagate throughout the hydrologic cycle.

In this study a generalized uncertainty distribution is introduced to model the probability distribution of the Stage IV Multi-sensor Precipitation Estimates (MPE) as the reference measurement given the PERSIANN satellite-based precipitation product (Precipitation estimation from remotely sensed information using Artificial Neural Network). The model is calibrated for an area of 5° × 5°, over the southeastern United States for both summer and winter seasons separately from 2004-2009. The uncertainty model parameters are further extended across various rainfall rates and spatial and temporal resolutions.

The method is evaluated for the period of 2006-2008 over the Illinois River watershed south of Siloam Springs, Arkansas. Results show that, using the proposed method, the estimation of the precipitation is improved in terms of percent bias and root mean squared error.

To further study the hydrological response of the satellite precipitation uncertainty; this uncertainty model is propagated as an input into the SAC-SMA (Sacramento Soil Moisture Accounting) hydrology model over the same case study watershed. The results shows that the proposed uncertainty model improves the simulated streamflow from the PERSIANN satellite precipitation product with regards to its, percent bias by more than 90 % and the root mean squared error by more than 30 %.

The uncertainty model is also applied to for the new GPM satellite precipitation product using the IMERG algorithm and results shows improvement in estimating the precipitation.