سال انتشار: ۱۳۹۰

محل انتشار: پنجمین کنفرانس بین المللی پیشرفتهای علوم و تکنولوژی

تعداد صفحات: ۹

نویسنده(ها):

A Gheiby – University of Hormozgan
R Rezvanizadeh –

چکیده:

Radiance measurements from satellites offer the opportunity to retrieve atmospheric variables at much higher spatial resolution than is presently afforded by in situ measurements (e.g., radiosondes). However, the accuracy of these retrievals is crucial to their usefulness, and the ill-posed nature of the problem precludes a straight forward solution. In this paper, an inversion neural network method has been investigated to retrieve Total Perceptible water (TPW), over Iran, from Advanced Microwave Sounding Unit-B (AMSU-B) measurements on NOAA-16 satellite. Because this satellite passes over Iran at approximately the radiosonde launch times, 0000 GMT, collocated radiosonde data were available for training and comparison with the satellite brightness temperature. The collocated radiosonde observations, at 00.00 GMT, and AMSU-B data during 2003 – ۲۰۰۷ are employed to build the neural network training and testing data sets. Overall 1250 days of collocated AMSU-B and radiosonde data were matched in this period. The Root Mean Square Error (RMSE) of TPW retrieved with neural network algorithm is about 3.45 mm and mean bias of 0.86 for entire sounding over Iran land. The RMS errors of the TPW retrieved with the trained neural network are compared with the errors from the multi-linear regression method. It is show that the neural network – based algorithm can provide much better results in the experimental region in all weather conditions