Please use this identifier to cite or link to this item:
|Title:||Using surface classification to improve surface reference technique performance over land|
|Authors:||Durden, S L|
Surface reference technique performance
|Abstract:||The surface reference technique (SRT) is the primary method of estimating atmospheric attenuation for spaceborne rain radars. The authors hypothesize that part of the error in applying the along-track spatial SRT over land is due to the difference in land type between the reference area and the rain area. Tropical Rainfall Measuring Mission Precipitation Radar (TRMM PR) data is used to show that dividing the data into classes (unsupervised clustering) reduces the standard deviation of surface backscatter, especially near nadir incidence. Similar clustering at Ka-band is also found for a small set of aircraft data. These results are applied to the SRT by creating a new, class-based SRT in which a reference is maintained for each land class. The new SRT is shown to have smaller difference between forward and backward runs (a measure of the error), as compared with the standard (one-class) SRT, especially near nadir.|
|ISSN:||0975-105X (Online); 0367-8393 (Print)|
|Appears in Collections:||IJRSP Vol.41(4) [August 2012]|
Items in NOPR are protected by copyright, with all rights reserved, unless otherwise indicated.