5.1 Conclusions
A diverse interaction of moist processes at the ocean surface, in the free atmosphere, and within clouds, is responsible for a large-scale cycling of moisture through the monsoon system providing diabatic heating to drive the monsoon circulation. Here we have used the most recent satellite and reanalysis datasets to diagnose these processes.

Some salient conclusions regarding the reanalyses include:

  • Moisture transports represented in the reanalyses agree closely (relative to most other hydrologic components). Principal boundary fluxes [those into the Arabian sea from the south, into India from the west, and out of the Bay of Bengal to the east] agree to within 10%.
  • Precipitation estimates show larger disagreement (>50 W m-2). ECMWF generally shows larger rainfall amounts. Though satellite estimates are unable to verify which reanalysis is correct in the monsoon region, the extended ITCZ over the western Indian Ocean in the NCEP/NCAR reanalysis appears to be inconsistent with available retrievals.
  • Disagreement in evaporation is generally on the order of 10-20 W m-2. The ECMWF reanalysis shows greater evaporation in the southern Indian Ocean (which serves as the principle source of atmospheric moisture divergence).


    • Satellite retrievals of the hydrologic cycle show that:

  • Estimates of evaporation from HOAPS are probably unrealistic in the monsoon environment. Problems in the retrieval process appear to be linked to the retrieval of low-level humidity.
  • Rainfall estimates from satellite, though useful in identifying a spatial distribution, are unable to specify the absolute magnitude to within 50%. The difference between the reanalyses is small than between GPCP GPI and multi-satellite products over much of the monsoon environment.
  • Large discrepancies are apparent between the reanalyses and satellite retrievals over land. Over ocean, only the NCEP/NCAR reanalysis deviates substantially from NVAP.


    • Model simulations of the hydrological cycle reveal that:

  • Large differences exist between the simulation of water vapor transports by the CCM3 and the reanalyses.
  • A distorted distribution of moisture convergence (P-E) in the CCM3 biased towards the Arabian Sea and away from southeast Asia.


    • Future work:

    Subsequent model studies will investigate the causal relationships between various components of the hydrologic cycle. The role of hydrologic variations in interannual variability of the monsoon system (talk 12C.3) will also be explored.


    [back to home][introduction][reanalysis][satellite][ccm3][conclusions]
    [to beginning]