Premium
Underestimation of deep convective cloud tops by thermal imagery
Author(s) -
Sherwood Steven C.,
Chae JungHyo,
Minnis Patrick,
McGill Matthew
Publication year - 2004
Publication title -
geophysical research letters
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 2.007
H-Index - 273
eISSN - 1944-8007
pISSN - 0094-8276
DOI - 10.1029/2004gl019699
Subject(s) - cloud height , cloud top , cirrus , lidar , tops , liquid water content , opacity , cloud computing , radiative transfer , environmental science , cloud albedo , cloud cover , brightness , albedo (alchemy) , thermal , optical depth , brightness temperature , meteorology , remote sensing , geology , physics , astronomy , optics , computer science , azimuth , operating system , art , aerosol , performance art , art history
The most common method of ascertaining cloud heights from space is from thermal brightness temperatures. Deep convective clouds of high water content are expected to radiate as black bodies. Here, thermal cloud top estimates from GOES‐8 are compared with direct estimates of where the top should be sensed, based on colocated Goddard Cloud Physics Lidar (CPL) observations collected during the Cirrus Regional Study of Tropical Anvils and Florida Area Cirrus Experiment (CRYSTAL‐FACE). GOES‐8 cloud top heights are consistently ∼1 km lower than the “visible” cloud top estimates from the lidar, even though the latter take into account the finite visible opacity of the clouds and any overlying thin cirrus layers, and are often far below the position of highest detected cloud. The low bias in thermal estimates appears to get worse for the tallest clouds, perhaps by an additional kilometer, and depends little on cloud albedo. The consistency of the bias over multiple satellites suggests that cloud retrievals are affected by an unexpected radiative transfer issue.