Open AccessOperational use of improved profiles by using neural network technique derived from NOAA satellites microwave data in NWP model over Indian region
Author(s) -
Devendra Singh,
Y. V. Rama Rao,
R. C. Bhatia,
S. K. Srivastav,
Sant Prasad,
Sanjoy Kumar Mukharjee
Publication year - 2005
Publication title -
mausam
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 0.243
H-Index - 12
ISSN - 0252-9416
DOI - 10.54302/mausam.v56i2.937
Subject(s) - artificial neural network , meteorology , data set , environmental science , interpolation (computer graphics) , remote sensing , data processing , computer science , geography , artificial intelligence , database , motion (physics)
India Meteorological Department, New Delhi receives and process NOAA TOVS and ATOVS data in real time. The physical and neural network approaches have been used to retrieve atmospheric temperature and moisture profiles from NOAA-16 & 17 satellites AMSU data over Indian region. The earlier training data set based on global data only for two seasons used in neural network technique has been replaced by new training data set based on regional data over land and ocean for all the seasons. The new training data set has improved the temperature and moisture profiles accuracy retrieved using neural network approach compared to physical method. The detail validation and inter comparisons of temperature and moisture profiles have also been carried out with ECMWF analysis over sea and land separately for different seasons for the year 2002-2003. The performance of neural network technique is found to be superior compared to physical method.
Recently, temperature and moisture profiles retrieved from NOAA-16 ATOVS data over Indian region have been used in regional NWP model for the impact study. The operational NWP system of India Meteorological Department is based on a Limited Area Analysis and Forecasting System (LAFS), which consists of real time processing of data received on Global Telecommunication System (GTS), objective analysis by 3-D multivariate optimum interpolation (OI) scheme and a multi-layer primitive equation model. Several experiments were performed using temperature and moisture profiles retrieved from NOAA-16 ATOVS data. Using this data several experiments were undertaken to examine the impact of these data sets on some of the important weather systems such as monsoon depression, active monsoon conditions during monsoon 2003. The preliminary studies reveal that these additional data have a positive impact on rainfall prediction of the limited area model. Results of specific cases of impact studies are presented in the paper.