|Li, L -|
|Mcwilliams, G -|
|Gaiser, P -|
Submitted to: Meeting Abstract
Publication Type: Abstract Only
Publication Acceptance Date: February 15, 2010
Publication Date: March 1, 2010
Citation: Li, L., Hunt, E.R., McWilliams, G., Gaiser, P. 2010. WindSat soil moisture and vegetation validation and performance prediction for the NPOESS microwave Imager/Sounder (MIS) [abstract]. 11th Specialist Meeting on Microwave Radiometry and Remote Sensing of Environment. 2010 CDROM. Technical Abstract: The National Polar-orbiting Operational Environmental Satellite System’s (NPOESS) Microwave Imager/Sounder (MIS) instrument is in development, with soil moisture sensing depth as one of the two Key Performance Parameters (KPPs). The other one is ocean surface wind speed precision. Based on the current design, the MIS sensor shares many channel configurations similar to the WindSat instrument, which provides an opportunity to predict MIS soil moisture performance using WindSat data. The WindSat land surface algorithm is a physically-based algorithm used to retrieve simultaneously the soil moisture, land surface temperature and vegetation water content for a range of surface types except for snow, frozen, rainy and flood surfaces. The algorithm is based on maximum-likelihood estimation and uses dual-polarization WindSat passive microwave data at 10, 18.7 and 37 GHz. The global soil moisture retrievals are validated at multi-spatial and multi-temporal scales against soil moisture climatologies, in situ network data, and precipitation patterns. The performance of the estimated volumetric soil moisture was within the requirements for most science and operational applications (standard error of 0.04 m3/m3, bias of 0.004 m3/m3, and correlation coefficient of 0.89). The comparisons between the WindSat vegetation retrievals and the MODIS vegetation water content show good agreement below 3 kg/m2, beyond which MODIS data saturate and do not respond to vegetation growth. In addition comparison WindSat vegetation water and AVHRR Green Vegetation Fraction data also reveals the consistency of these two independent data sets in terms of spatial and temporal variations. The algorithm has also demonstrated great science valuel in study of soil moisture response to ITCZ (Intertropical Convergence Zone) propagation, drought detection, and heat wave evolution. The evaluation results suggest that the WindSat data products meet IORD II threshold soil moisture requirements. To approximate MIS performance, a land algorithm was adapted from WindSat to run on WindSat proxy data, which simulate MIS SDRs by adding Gaussian noise to WindSat SDRs according to their differences in effective NEDTs. MIS effective NEDTS are derived from 45 km CFOV (Composite Field of View) sizes that meet soil moisture HCS requirements. This proxy data approach cannot accommodate the EIA (Earth-Incidence-Angle) increase from WindSat to MIS; but a simplified analytical estimation indicates that the impact of such an EIA increase on soil moisture performance varies from low, for bare soil, to moderate, for dense vegetation. The simulation results suggest that the soil moisture performance meets the MIS algorithm specification with 40% margin given that the RFI impacts can be effectively mitigated. There is virtually no difference between MIS and WindSat performance in soil moisture retrieval based on the current MIS design.