Abstract:
Numerical model output's verification and interpretation are very important to m
odel's development and maximum operational application. Shanghai Regional Meteor
ological Center (SRMC)'s regional numerical model has been in operation for near
ly 10 years. Evaluation on its forecasting capabilities must be conducted for it
s scientific developing. On the other side, model output interpretation has been
entering a new time of consensus theory. Applications of numerical model output interpretation based on Kalman filter and optimal consensus foreca
st are introduced and verified in SRMC.
After detailed verification and comparison on Direct Model Output (DMO),Kalman f
ilter interpretation(KLM),Optimal Consensus Forecast(OCF) for SRMC's regional nu
merical model, it is found that (1) For temperature, relative humidity and wind
, the performance of SRMC's DMO does not show a trend of improvement, but a litt
le declination, though it has had a more stable wind speed forecast since 2005.
(2) The comparison of KLM and DMO shows that the forecast stability of KLM behav
es much higher than that of DMO in temperature, relative humidity and wind speed
. But its forecast on wind direction shows no advance with a slightly lower accu
racy than DMO. (3) OCF performs better than KLM. For temperature, relative humid
ity and wind direction's accuracies, it shows an improvement of 2%. As to wind s
peed, though its accuracy is nearly equal to KLM's, it has a lower average error
than the latter. In the spring and winter, the OCF stands at the same accuracy
level with official forecast. With the guidance of OCF, the weakness of official
forecast, “Which is error in the spring and winter than other seasons”, can b
e mended. The verification results could be good references for numerical model'
s development and transplantation. The success of OCF in SRMC's routine operatio
n projects a bright prosperity for the application of consensus theory on numeri
cal model objective interpretation.