Influence of Ground Observation Stations in Merging Satellite Rainfall Data with Ground Observation Data
Abstract
The complex topography with many mountainous regions greatly affects the spatial rainfall distribution. However, there is a limitation in reproducing the spatial distribution of actual rainfall from the existing ground observation network due to its uneven spatial distribution. Therefore, if the spatial rainfall distribution of the watershed is estimated based on the ground observation network, it may be underestimated or overestimated, which causes errors in evaluating the hydrological cycle of the watershed. Satellite rainfall data is advantageous to estimating the availability of water resources in mountainous regions where ground observations are rare. Global Satellite Mapping of Precipitation (GSMaP), which provides global hourly rainfall data with a spatial resolution of 0.1 degrees, is widely used not only for scientific and educational purposes but also for disaster prevention, climate monitoring, and water resources management.
This study used a conditional merging technique to merge the ground observations with GSMaP for the Namgang Dam watershed in South Korea. To separate rainfall events in a year, the IETD (InterEvent Time Definition) concept was applied to the ground hourly rainfall time series of the past 10 years (2012 to 2021) at a total of 55 rainfall stations where the 39 rainfall events (at least one event per year) were selected. A bilinear interpolation technique was used to downscale the GSMaP data and the conditional merging technique was applied to bias correction. The co-kriging technique was applied to interpolating ground rainfall data to consider the effects of terrain elevations. A leave-one-out cross-validation (LOOCV) concept was applied to validating the accuracy the merging technique in simulating the cumulative rainfall depth. The Nash-Sutcliffe Efficiency (NSE) was used as a performance indicator to represent the influence (or importance) of each station included in the merging technique. The proposed merging technique will be useful to provide a value of information of each station (i.e., observation location) for designing an optimal observation network and simulating a rainfall-runoff model. Acknowledgments This work was supported by the National Research Foundation of Korea (NRF) grant funded by the Korea government (MSIT) (No. NRF-2022R1A2B5B01001750).- Publication:
-
AGU Fall Meeting Abstracts
- Pub Date:
- December 2022
- Bibcode:
- 2022AGUFM.H35P1320L