Verification of the TIGGE and WRF Models in forecasting the precipitation event for 3rd February 2010 over Zambia
In this case study, attempt was made to verify the performances of the TIGGE and WRF models in simulating rainfall over southern Africa for the period from 1st November 2009 to 10th February 2010, with 24 and 48 hrs lead time. We have tested the skills of the models in forecasting heavy rainfall event, in excess of 40mm per day. Special emphasis was given to rainfall event that occurred on the 3rd of February 2010 over Zambia. The associated atmospheric conditions prior and during the rainfall event were also analyzed, both in analysis and forecasted fields.
Over Southern Africa, all the models followed the general pattern of the observed rainfall but missed extreme rainfall events, with the BOM model under forecasting throughout the period. The overall statistical verification methods indicate good performance for the ECWMF model over southern Africa while showing that the BOM model underperformed.
For the 3rd February rainfall events over Zambia, the CPC/RFE performed well in estimating rainfall amount and distribution over Zambia. The model evaluation for the rainfall event of 3rd February, 2010 indicates the NCEP model performed better in capturing the pattern of events as compared with the other TIGGE models. The NCEP model had minimum errors overall and showed a better correlation between the observed and forecasted rainfall over Zambia. The NCEP model consistently performed well with all the statistical verification tests carried out over Zambia.
Notes: If you wish to download the full text, please contact library(at)wmo.int (Please replace (at) by @).
Format: Digital, Hard copy (ill., charts, maps)