الفهرس 
Rainfall Estimation Challenges (Pros and Cons)
MethodologyOnly 14 pages are availabe for public view
 |  | from | 137 |  |  |
| | | from | 137 | | |
Abstract
Ground precipitation measurements face obstacles and limitations in many regions in the world, where gauge stations are sparse or even nonexistent. Thus, it is necessary to find reliable sources such as satellite-based precipitation products which provide uninterrupted precipitation time series with high spatial global coverage. Yet, the hydrological applications of satellite-based precipitation data are rather limited due to its biased estimation. Thus, the purpose of this work is to evaluate the performances of three well-known global satellite precipitation products: Precipitation Estimation from Remotely Sensed Information using Artificial Neural Networks-Climate Data Record (PERSIANN-CDR), Tropical Rainfall Measuring Mission Multi-Satellite Precipitation Analysis 3B42 product (TRMM3B42V7), and the gauge corrected Global Satellite Mapping of Precipitation (GSMaPV6) against daily gauged data of 23 stations in Egypt for the period of (2003-2014) at daily and annual scale. Furthermore, the most consistent satellite product will be corrected by three commonly used methods: linear scaling (LS), local intensity scaling (LOCI), and empirical quantile mapping (EQM). Accuracy in estimating rainfall amounts and the ability in detecting rainfall occurrences were assessed through statistical continuous indices, frequency-based indices, and categorical indices. The results showed that there were significant biases in raw outputs produced by the three products. However, GSMaPV6 revealed outstanding skill over the other two products in detecting rainfall occurrences and estimating the amount of rainfall on a daily scale. On the other hand, both TRMM3B42V7 and PERSIANN-CDR become more skillful as the time scale increases. Overall, the results showed that GSMaPV6 is best suited for hydrological applications in Egypt, however, preprocessing of its estimates is a perquisite step to reduce errors. All three bias correction methods are effective in