Share Email Print

Proceedings Paper

Statistical properties of intercalibration techniques that use empirical distribution functions
Author(s): David S. Crosby; Elena Randou
Format Member Price Non-Member Price
PDF $14.40 $18.00
cover GOOD NEWS! Your organization subscribes to the SPIE Digital Library. You may be able to download this paper for free. Check Access

Paper Abstract

Techniques which use empirical distribution functions have been used successfully to intercalibrate detectors on the same instrument or channels on different instruments. In this paper the procedure is developed as a statistical technique for estimating a function. Since it is a statistical procedure, it is possible to use classical statistical techniques to evaluate the accuracy of the results. Confidence regions can be constructed for the estimated parameters. Given a required degree of accuracy it is possible to estimate the required sample sizes. Using hypotheses testing techniques questions about possible changes in the inter-calibration functions can be answered. The theory, which uses quantiles from the distributions of random variable, is outlined. The large-sample properties of the estimators of these quantiles are presented. The theory requires that the data be continuous. However, it is discrete. Preliminary simulation results indicate that the theory for continuous data can be applied to the discrete data with little loss of accuracy.

Paper Details

Date Published: 18 October 1996
PDF: 4 pages
Proc. SPIE 2812, GOES-8 and Beyond, (18 October 1996); doi: 10.1117/12.254106
Show Author Affiliations
David S. Crosby, National Environmental Satellite Data and Information Service/NOAA (United States)
Elena Randou, American Univ. (United States)

Published in SPIE Proceedings Vol. 2812:
GOES-8 and Beyond
Edward R. Washwell, Editor(s)

© SPIE. Terms of Use
Back to Top