In this paper, an artificial neural network (ANN) technique is developed to obtain vertical profiles of ... cyclone category that complements the Dvorak technique.
DOI https://doi.org/10.1007/978-94-007-7720-0_34 Publisher Name Springer, Dordrecht Print ISBN978-94-007-7719-4 Online ISBN978-94-0077720-0 ,eBook PackagesEarth and Environmental Science
Retrieval of atmospheric temperature profiles from AMSU-A measurement using artificial neural network and its applications for estimating tropical cyclone intensity A.K. MITRA1 , A.K. SHARMA1 ,P.K. KUNDU2 1 2
India Meteorological Department, New Delhi
Department of Mathematics, Jadavpur University, Kolkata-700032, India Abstract
In this paper, an artificial neural network (ANN) technique is developed to obtain vertical profiles of temperature from Advanced Microwave Sounding Unit-A (AMSU-A) measurements over the Indian region. The datasets, in the form of level 1b (instrument counts, navigation and calibration information appended) format pre-processed by ATOVS (Advanced TIROS Operational Vertical Sounder) and (AVHRR) Advanced Very High Resolution Radiometer Processing Package (AAPP). The corresponding global analysis data generated by National Center for Environmental Prediction (NCEP) and AMSU-A data from April 2007 to March 2008 are used to build the neural network training data-sets and from April 2008 to July 2008 used as independent dataset. The root mean square (RMS) error of temperature profile retrieved with the ANN is compared with the errors from the International Advanced TOVS (ATOVS) Processing Package (IAPP). The overall RMS errors of ANN are found to be less than 3°C at the surface, 0.9° to 2.8° between 700-300 hPa and less than 2°C between 300-100 hPa. The comparison has also been carried out using radiosonde temperature profiles for independent datasets to test the quality of temperature profile retrievals from ANN and IAPP. It has been observed that the neural network technique can yield remarkably better results than IAPP at the low levels and at about 200-hPa level. The highlight of the work is however, the case studies of upper tropospheric warm core anomaly of two cyclones. To critically examine the performance of the retrievals from ANN for an extreme event, the network-based retrieval algorithm applied for computing the anomaly near the center of the cyclone 'gonu' and 'nargis', formed over Arabian Sea (May-June 2007) and Bay of Bengal (April-May 2008) respectively. Further, the anomalies are related to the intensification of the cyclone. It has been found that the temperature anomaly at 200 hPa can be a good indicator of the intensity of tropical cyclone and as cyclone intensifies, the warm core temperature anomaly increases with different intensities of cyclone, which is the indication of the positive relationship of cyclone intensities with warm core anomaly. Therefore it may be stated that optimized neural network can be easily applied to AMSU-A retrieval operationally and it can also offer substantial opportunities for improvement in tropical cyclone studies and appears promising for analyzing the cyclone category that complements the Dvorak technique. Keywords: ANN, AAPP, IAPP, AMSU-MBT, brightness temperature (TB), NOAA, AMSU
