Sensitivity of Single-Year Seasonal Precipitation to Parameterization in the Weather Research and Forecasting (Wrf) Model

Abstract

The production of precipitation in a regional climate model (RCM) is handled primarily by the microphysics and cumulus physics parameterizations. Application of the WRF model as an RCM to a midlatitude Japan case and tropical Philippine case using single-year simulation sensitivity experiment showed scheme sensitivity to simulation of precipitation and exploration of potential transferability of the schemes to different climate regime applications. Results show that simulations can capture seasonal 2-meter air temperature well and demonstrate capability in simulating seasonal precipitation for both application. Results show that cumulus (microphysics) scheme sensitivity is higher in simulating seasonal and monthly precipitation when larger fraction of precipitation is convective (non-convective). Precipitation sensitivity to cumulus scheme showed largest variation in warm season for Japan case and in all seasons in Philippine case. It has been found that Kain-Fritsch (KF) scheme tend to overestimate precipitation in hot humid climate regime application and Grell-3 (GR) scheme showed the most consistent good skill performance throughout the seasons for both applications. CU scheme selection showed higher impact for simulating precipitation in the model than the MP scheme selection.

Evaluation of WRF Parameterizations in Simulating Seasonal Precipitation Over Japan

“…Results for CU sensitivity experiment were already discussed in previous paper 5) focused on selection of CU-MP combination of schemes for Japan and Philippines. This paper, however, focuses on the CU scheme coupling with MP schemes, RAD schemes, and BL schemes in representing precipitation in the WRF simulations through PCC-SDR plots and mean budget profile analysis.…”

“…On the other hand, wide variation of the results for JJA and SON seasons. This shows the control of small-scale forcing in the precipitation mechanism during these seasons defined by the occurrence of Baiu front, extratropical typhoons, and thunderstorms 5) . This variation show the sensitivity of seasonal precipitation to precipitation forcing as represented by convection triggers in the cumulus convection schemes.…”

“…However, an issue was also documented regarding an anomalous negative precipitation bias during months of October, November, and December that is accompanied by positive 2-meter temperature biases 5) . Due to the said issues, additional validation of the vertical structure of the simulations were done using AIRS satellite data for several pressure levels (850hPa, 700hPa, 500hPa, 300hPa, 200hPa).…”

“…This study investigated the selection of cumulus schemes and its combination with other parameterizations such as microphysics schemes, radiation schemes, and boundary layer schemes in the Weather Research and Forecast (WRF) model in reproducing seasonal precipitation over mid-latitude Japan by utilizing a 24-km one-year hindcast simulation from June 2005 to May 2006. The selection of cumulus and microphysics schemes had already been published in a previous paper by the authors focusing on Taylor diagram, mean bias analysis, and physical consistency of schemes based on underlying scheme development assumptions 5) . However, this paper supplements the previous study 5) and focuses on the usage of domain-time mean budget analysis in diagnosing the impact of using cumulus scheme and its combination to the other parameterizations in reproducing the apparent heating/drying processes related to production of precipitation in the model.…”

“…The selection of cumulus and microphysics schemes had already been published in a previous paper by the authors focusing on Taylor diagram, mean bias analysis, and physical consistency of schemes based on underlying scheme development assumptions 5) . However, this paper supplements the previous study 5) and focuses on the usage of domain-time mean budget analysis in diagnosing the impact of using cumulus scheme and its combination to the other parameterizations in reproducing the apparent heating/drying processes related to production of precipitation in the model. Validation of the simulations were mainly done against daily gridded precipitation data from the APHRODITE's Water Resources products, which are available from http://www.chikyu.ac.jp/precip/products/index.html.…”