Several strategies for establishing a land surface emissivity scheme for AMSU have been tested at the Met Office in recent years. The initial approach involved the use of an atlas to give a first guess estimate of emissivity, and a subsequent test involved retrieving inputs to the FASTEM emissivity parameterization (Hewison and English, 1999) as part of a 1D-Var retrieval. Neither of these approaches were particularly successful, producing little benefit in observed minus background (O-B) quantities, and the latter being hampered by poor convergence rates. A new approach is under development to use Weng and Yan’s (2003) microwave snow emissivity model to provide a first guess emissivity where appropriate and to retrieve emissivities directly in the 1D-Var. The reduced O-B values from this scheme are a promising step.
[1]
Catherine Prigent,et al.
Microwave land surface emissivities estimated from SSM/I observations
,
1997
.
[2]
Catherine Prigent,et al.
Microwave land emissivity calculations using AMSU measurements
,
2005,
IEEE Transactions on Geoscience and Remote Sensing.
[3]
Fuzhong Weng,et al.
A Microwave Snow Emissivity Model
,
2003
.
[4]
Tim J. Hewison,et al.
Airborne retrievals of snow and ice surface emissivity at millimeter wavelengths
,
1999,
IEEE Trans. Geosci. Remote. Sens..
[5]
Ralf Bennartz,et al.
Optimal Convolution of AMSU-B to AMSU-A
,
2000
.