Response of the WRF model to different resolutions in the rainfall forecast over the complex Peruvian orography

Aldo S. Moya-Álvarez, Daniel Martínez-Castro, Shailendra Kumar, René Estevan, Yamina Silva

Research output: Contribution to journalArticlepeer-review

34 Scopus citations

Abstract

The main objective of the research is to evaluate the response of the WRF model to the domains and resolutions that are used in complex orographic conditions like the central Andes of Peru for the forecast of short- and medium-term rainfall. To do this, the model was configured with four domains and the verifications were made using data from meteorological stations located within the study area and TRMM data. Experiments were conducted for nine 10-day periods of rainy days, five cases of extreme rainfall, and one event with hail fall on the region. In general, the model overestimates precipitation, but, in the five cases of extreme rainfall, and in the case of the hailstorm, underestimation was observed, so it is not accurate to assert in an absolute way that WRF overestimates precipitation in the study region. It was observed that the 3-km domain simulate effectively the accumulated rainfall, while the 0.75-km domain reproduces better the process at local scale. The results in the domain with the coarsest resolution of 18 km showed the lowest skill in simulating rainfall compared to the higher resolutions. Thus, it is concluded that an increase of resolution leads to an improvement of the results of rainfall forecast in the region and the structure of clouds systems. At the same time, the domains with resolutions of 18 km showed poorer results.

Original languageEnglish
Pages (from-to)2993-3007
Number of pages15
JournalTheoretical and Applied Climatology
Volume137
Issue number3-4
DOIs
StatePublished - 1 Aug 2019
Externally publishedYes

Fingerprint

Dive into the research topics of 'Response of the WRF model to different resolutions in the rainfall forecast over the complex Peruvian orography'. Together they form a unique fingerprint.

Cite this