Open AccessUncertainty Characteristics of Total Water Path Retrievals in Shallow Cumulus Derived from Spaceborne Radar/Radiometer Integral Constraints
Author(s) -
Matthew Lebsock,
Kentaroh Suzuki
Publication year - 2016
Publication title -
journal of atmospheric and oceanic technology
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 0.774
H-Index - 124
eISSN - 1520-0426
pISSN - 0739-0572
DOI - 10.1175/jtech-d-16-0023.1
Subject(s) - environmental science , remote sensing , liquid water path , brightness temperature , radiometer , microwave radiometer , attenuation , water vapor , microwave , radar , meteorology , brightness , physics , geology , optics , computer science , aerosol , quantum mechanics , telecommunications
A precipitating marine cumulus cloud simulation is coupled to radiation propagation models to simulate active and passive microwave observations at 94 GHz. The simulations are used to examine the error characteristics of the total water path retrieved from the integral constraints of the passive microwave brightness temperature or the path-integrated attenuation (PIA) using a spatial interpolation technique. Three sources of bias are considered: 1) the misdetection of cloudy pixels as clear, 2) the systematic differences in the column water vapor between cloudy and clear skies, and 3) the nonuniform beamfilling effects on the observables. The first two sources result in biases on the order of 5–10 g m −2 of opposite signs that tend to cancel. The third source results in a bias that increases monotonically with the water path that approaches 50%. Nonuniform beamfilling is sensitive to footprint size. Random error results from both instrument measurement precision and the natural variability in the relationship between the water path and the observables. Random errors for the retrievals using the CloudSat PIA are estimated to be the larger of either 20 g m −2 or 30%. A radar/radiometer system with a measurement precision of 0.3 K or 0.05 dB could reduce this error to the larger of either 10 g m −2 or 30%. All error mechanisms reported here result from variability in either the spatial structure of the atmosphere or the hydrometeor drop size distribution. The results presented here are specific to the cloud simulation and in general the magnitude will vary globally.