A Single Field of View Method for Retrieving Tropospheric Temperature Profiles from Cloud-Contaminated Radiance Data

摘要

A method is presented to retrieve single field of view (FOV) tropospheric temperature profiles directly from cloud-contaminated radiance data through the use of auxiliary data such as observed shelter temperature and estimated cloud-top height The iterative technique utilized is an extension of the work of Chabine as modified by Smith and Duncan It is shown that a well-defined temperature profile may be calculated from the radiative transfer equation (RTE) for a partly cloudy atmosphere when the average fractional cloud amount and cloud-top height for the FOV are known A cloud model is formulated to calculate the fractional cloud amount from an estimated cloud-top heightnThe method is then examined through use of simulated radiance data calculated through vertical integration of the RTE for a partly cloudy atmosphere using known values of cloud-top height(s) and fractional cloud amount(s) Temperature profiles are retrieved from the simulated data assuming various errors in the cloud parametersnTemperature profiles are retrieved from NOAA-4 satellite-measured radiance data obtained over an area dominated by an active cold front and with considerable cloud cover Excellent radiosonde data for 11 stations participating in the Atmospheric Variability Experiment (AVE III) available for the area are used for comparison with temperature profiles retrieved from the NOAA-4 data The effects of using various guessed profiles and the number of iterations are considerednThe results of the investigation indicaten(1) The single FOV method presented improves the accuracy of guessed profiles when retrievals from cloud-contaminated radiance data are attempted It is significant that in the method presented improvement innthe guessed temperature profile was noted under the cloud layer where retrievals using other single FOV techniques tend to deterioraten(2) Through proper choice of a guessed profile it appears possible to improve many profiles independent of the cloud amount presentn(3) Use of the method presented should provide useful data for mesometeorological research in the form of an accurate temperature profile representative of a relatively small horizontal area or many temperature profiles to show the pattern of temperature change over a larger area However the absolute accuracy of the retrieved profile is also a function of the a priori knowledge of the state of the atmosphere possessed by the researchern(4) It is possible to make an accurate estimate of the average tops of a thick overcast layer through use of ] the cloud model when there are no thick clouds present above the overcast layern(5) Observed cloud parameters do not possess sufficient precision for use directly in the RTE to retrieve temperature profiles.