ISSN 1000-0526
CN 11-2282/P
CN 11-2282/P
Verification of GRAPES-Meso Precipitation Forecasts Based on FineMesh and Station Datasets
- Article
- Figures
- Metrics
- Preview PDF
- Reference
- Related
- Cited by
- Materials
Abstract:
The highdensity precipitation data from both meteorological and hydrological observations in China are interpolated to finemesh grid at a resolution of 0.15° by combining the Barnes scheme with precipitation frequency, then 0-24 h quantitative precipitation forecast (QPF) from GRAPES_Meso is interpolated to the station spot. The QPF has been verified against two datasets using BIA, ETS and TSS. Moreover, an objectbased verification developed recently is carried out to verify the raincluster more than 5 mm. The results are as follows: (1) For all thresholds, the QPF exhibits a higher performance in the east part of China than that in the west, as well as more skillful in plain areas than in mountainous ones. The model forecasts more events above the threshold than observed in the upwind side of mountains (overforecasting) while the model forecasts less events than observations for interior Tibet and the north of Xinjiang (underforecasting). (2) As a whole, the scores calculated for the verifications against gridded observations have shown that the model performs better than could be anticipated from local observations, thus gridded observations better represent the gridbox behavior described by the model. (3) For the threshold more than 5 mm, the distribution of number against the cluster size of the QPF is similar to observations. Especially, the best forecast is to the cluster size of 60 km. As for spatial distributions, in the east of Inner Mongolia, the Northeast China, North China, the region between the Yangtze River and Huaihe River, Shaanxi Province as well as the south of Gansu Province, numbers of QPF are almost the same as observed ones, but more to the south of Yangtze River, especially in the east part of Southwest China.
Keywords:
Project Supported:
Clc Number:
Mobile website
