연구보고서
- 저자
- 박경원 박사
- 작성일
- 2019.06.13
- 조회
- 227
- 요약
- 목차
Research-based countermeasures to respond preemptively to disasters caused by climate change must be established for various sectors. For this, it is necessary to collect past and present weather monitoring data and pre-process it using other reliable data. Many advanced countries, including Korea, have built considerable climate monitoring systems and utilize the data obtained at the branch level. However, most of these climate monitoring systems in the Asia-Pacific region have not been constructed systematically. Considering this, satellite image-based climate data can provide an alternative to the absence of data for the Asia-Pacific region. Furthermore, there is an increasing need for grid-based climatic data in various application sectors. In this study, we developed a climate management system that produces grid-based precipitation data from satellite imagery to provide reliable climate data in the Asia-Pacific region. This system provides three kinds of satellite data (raw, bias-corrected, and spatially downscaled). The spatial resolution of the satellite data is 0.1° and 0.05° and the temporal resolution is 1 day. The Korean peninsula was selected as a target area, the original satellite data was constructed, and the correction method and spatial specification method were validated. We applied the technique tested and proved on the Korean Peninsula to Vanuatu to produce the final satellite data. The raw satellite data was constructed using the Tropical Rainfall Measurement Mission (TRMM) and Global Precipitation Mission (GPM) satellites. The Geographical Ratio Analysis-Inverse Distance Weighted (GRA-IDW), GRA-Kriging, and Quantile Mapping (QM) methods were developed to correct bias and GRA-IDW was finally selected. Spatial downscaling was performed using Parameter-elevation Regressions on Independent Slopes Model (PRISM). The correlation coefficient between 1998 and 2017, 0.775, is relatively high; and the 0.1° daily rainfall correlation coefficients for the Multi-satellite Precipitation Analysis (TMPA) and the Integrated Multi-satellitE Retrievals for GPM (IMERG) are 0.776 and 0.753, respectively. The bias value presents that raw satellite rainfall is overestimated than observed rainfall from Automated Synoptic Observing System (ASOS). As a result of cross-validation of the bias-correction methods, GRA-IDW method shows better correction results than GRA-Kriging and QM based on the correlation coefficient and Root Mean Square Error (RMSE). The results of spatial downscaling show that the correlation coefficient is smaller and RMSE is larger than that before spatial downscaling. However, since the difference is not large, it can be used in application sectors. It is possible to obtain reliable climate observation data to an ungauged watershed based on the technique developed in this study. It is also basic data for projects overseas, including in the Asia-Pacific region. It is expected that applying the system developed in this study to the Asia-Pacific region can provide reliable climate data on the latter.