Impact Of Satellite Data On Numerical Weather Prediction Modelling

A global spectral model (T-80/L/16) and a global data assimilation system based on short range forecast (six hours) of this model involving Spectral Statistical Interpolation (SSI) scheme of analysis has been operational at the NCMRWF since June 1994. The satellite data sets received on GTS along with many other data sets received directly from the satellite data sources are being routinely used on operational basis. Mitra et al. (2003) reported encouraging results made towards assimilation of different types of satellite data on the analysis and forecast system of NCMRWF. Inclusion of high density satellite winds improved the performance of the model in refining many of the flow features during the southwest monsoon season over India.

IMD is also running a limited area analysis and forecast system in which a variety of conventional as well as non-conventional data received on GTS system of WMO is being ingested. Prasad (1997) described the synthetic vortex generation scheme for numerical forecasting of TCs. The scheme basically generates radial distribution of surface pressure within the vortex from a empirical formula proposed by Holland (1980). Basic inputs for generating the surface pressure are the parameters like central pressure of the storm, its environmental pressure, radius of maximum wind, current position, movement and intensity of the storm, which are inferred from the satellite imagery. Surface winds are obtained from the gradient wind relationship. Upper winds are computed from surface winds by using composite vertical wind shear factors proposed by Anderson and Hollingsworth (1988). Recently a Quasi-Lagrangian Model (QLM) has also been installed. Inputs from the satellite data are quite important for initializing the vortex position.

Three areas wherein satellite data have significantly contributed to numerical weather prediction are: defining the initial conditions of the model, setting of the boundary conditions, and defining the forcing functions. The satellite derived parameters on SST, sea surface winds, temperature/humidity profiles etc. have been assimilated into the models and found to have significant impact on the forecast outputs. Boundary conditions play a crucial role in extended/seasonal/long-term predictions and several inputs such as SST-snow cover, vegetation cover, soil moisture, etc. are provided by satellite data. The impact studies carried out using the Extended Range Monsoon Prediction model (Pal et al., 1999) using SST and soil moisture has given new insights into their crucial role. One of the important forcing functions namely radiation budget operationally available from satellites is an important input to models. The most current research is focussing on assimilation of satellite inputs to models for improved performance.

Was this article helpful?

0 0

Post a comment