Comparing CGCM3 model output to observed divergence and temperature anomaly profiles

Convective clouds, in GCMs are sub-grid phenomena and therefore must be parameterized. Version 3 of the Canadian Coupled Global Climate Model (CGCM3) applies the Zhang-McFarlane (ZM) scheme to parameterize deep convection. We compare divergence profiles estimated from different rawinsonde rings to the divergence calculated from the models wind output. The comparison is done for both rainy and dry seasons. The modeled rainy season divergence is similar to the observations, while there is bias between modeled and observed dry season divergence. This bias may be a result of the absence of shallow convection or mesoscale downdrafts in CGCM3. We analyze the effect of evaporatively driven downdrafts on the observed temperature field in the regions with enhanced precipitation. The lower tropospheric cooling has an impact on the propagation of deep convection and consequently rainfall prediction. The results show that CGCM3 does not produce negative temperature anomalies in the regions of enhanced precipitation.