연구보고서
Development of APCC-PAGASA Regional Prediction System (APCC-PRePS) over the Philippines: Pilot Study
- 저자
- 양유빈 박사
- 작성일
- 2017.07.04
- 조회
- 245
- 요약
- 목차
This study is a part of the APCC-PAGASA Regional Prediction System (APCC-PRePS) project aimed at providing valuable high-resolution seasonal forecasts, using downscaling techniques. The purpose of this study, as a pilot study, is to test model setup and feasibility before inter-comparison of regional climate models.
Dynamical downscaling was carried out by the WRF, RegCM, and GRIMs-RMP with the same horizontal resolution and domain for May-June-July-August 1998. The results were evaluated by Tropical Rainfall Measuring Mission (TRMM) Multi-Satellite Precipitation Analysis (TMPA) in order to focus on simulated precipitation. This pilot study was done to address some issues with the operational
use of the regional climate model. As part of the research strategy, the following issues are discussed prior to the main study:
(1) Dynamical core
The results from parallel runs using the WRF and RegCM showed that the differences between the hydrostatic and non-hydrostatic model appear mostly over mountainous areas, even at a resolution of 30 km. We have to decide whether or not we have to pay extra computation resources to run the non-hydrostatic model for certain resolutions.
(2) Domain size
The results showed that monthly precipitation and daily evolution in terms of spatial and temporal correlation generally improve when the domain size is reduced. It is well known that the limited area over which a model is integrated must be large enough to allow for the full development of small-scale features. On the other hand, simulations in small domains with strong orographic forcing or land-sea contrast may enhance the accuracy of a regional climate model, and correct some errors at small scales.
(3) Physical options
The use of three convection schemes and their combinations showed that regional climate simulations are very sensitive to the physical parameterization schemes employed. An appropriate convective parameterization scheme should be used in dynamic models for realistic simulations.
(4) Spectral nudging
As an alternate method to mitigate systematic large-scale errors, the technique of large-scale spectral nudging was applied. If simulations over a large domain or for a period longer than one month are accessible, we may consider adapting a large-scale nudging approach in the regional climate model.