Abstract:
This research work evaluated the performances of five cumulus parametrization schemes (CPSs)
in Weather Research and Forecasting (WRF) model for simulating the initiation, propagation and
rainfall delivery capacity of mesoscale convective systems (MCSs) over West Africa. Five
simulations were conducted using five different cumulus schemes, for the whole of
September, 2012 with the initial and lateral boundary conditions from ERA-Interim reanalysis
data. The five cumulus parameterization schemes used are the Kain-Fritsch (KF), Grell-
3Dimension ensemble (Grell 3D), Tiedtke (TKE), Grell-Freitas (GF), and New Simplified
Arakawa-Schubert (NSAS). Infra-red brightness temperature from the satellite imagery was used
for the performance evaluation while TRMM rain rate were used as observed to compare rain rate
of the five CPSs. All the schemes successfully simulated spatial feature of MCS in the case 1,
although there was 2 hours delay/lag for Grell 3D, GF and NSAS schemes before they initiated
MCS. During initiation, TKE scheme performed better than the other schemes in reproducing
initiation of MCS in all the cases. The monitored MCS in case 1, case 2 and case 4 propagated
westward while it moved north-west in case 3. This may be caused by cold air-mass meeting with
MCS around Democratic Republic of Congo. Furthermore, the NSAS scheme gave better average
of daily rain-rate than the other schemes in all the cases, except Case 4 where KF performed better
than the other schemes with their smallest mean bias. It was discovered that the scheme which
performed better for average aggregate of daily rain-rate was almost the worst for hourly maximum
value of rain-rate. This may be due to local forces. However, based on the subjective verification
and objective evaluation, it can be concluded that the performance of the five schemes is highly
case-and local-dependent, and the performance of the better TKE and NSAS schemes cannot be
generalised over West Africa.
