3 rd JCSDA Science Workshop Assimilation of Infrared and Microwave Window Channels over Land Benjamin Ruston and Nancy L. Baker In collaboration with: Drs. Fuzhong Weng, Banghua Yan, and Andrew Jones NRL – Monterey 20 April, 2005 Assimilation of Passive Microwave Radiances over Land: Use of the JCSDA Common Microwave Emissivity Model (MEM) in Complex Terrain Regions
GOAL: - improve atmos profiling & radiance assimilation over land - improve the Land Surface Temperature (LST) - reduce rejects over desert & elevated terrain ISSUES: - true background land emission (validation) - uncertainty in land surface temperature analysis - behavior of surface emission characteristics (scales) How do the emission characteristics vary ? canopy properties soil properties surface roughness effectsMotivation
Milestones of the Past Year Began March 15, 2004 funding began Apr??, Gather vegetation & soil databases (GLDAS) - Develop IR emissivity & single-channel LST retrieval - Implement MEM into NAVDAS framework - Perform initial 1dvar emissivity retrieval -Combine infrared & microwave into single retrieval -Transition 1dvar for NRL beta-testing - Begin examination of bias correction
initially: IR single-channel LST retrieval retrieve emissivity statistics currently: AMSU/A and B emissivities correlated HIRS/3 emissivities independent all retrieved simultaneously 1DVAR Retrieval 1DVAR retrieval is used to handle non-linear surface sensing channels 1dvar preprocessor passes variables: sfc emissivity, LST
Quality Control redevelop HIRS/3 radiance quality control using simple cloud mask
Case Study Study periods: Mar-Apr 2003 & 04 Aug-Sep 2003 & 04Features: Changing season UAE 2 overlap Dust Storm Events
Infrared: Land Databases indexing to spectral library Surface Emissivity How is the first guess emissivity made? Microwave: 1dvar retrievals create statistical database MEM used for snow/ice & vegetation fractions > 75%
Emissivity Means 1-dvar retrieval incorporates HIRS/3, AMSU/A and AMSU/B to simultaneously retrieve profiles T & q, emissivity, and LST
Emissivity Std Deviation standard deviations small for ocean, highly vegetated land standard deviations larger for when intermittent snow/ice, coastlines
Transmission infrared “window” becoming opaque in tropics 183.3±7 GHz emissivity retrieval often relies on cross correlations
Future Work Coming Year: - complete CRTM implementation - transfer methodology to NCEP - mesoscale impact studies (UAE 2 ) - extend to SSMI/S assimilation - complete strategy for bias correction May need some help: - footprint matching (beam filling) - land surface model (Noah implementation/Teddy Holt)
Assimilation of Passive Microwave Radiances over Land: Use of the JCSDA Common Microwave Emissivity Model (MEM) in Complex Terrain Regions Dr. Nancy Baker (PI) & Benjamin Ruston (co-Investigator) Naval Research Laboratory, Monterey, CA collaborators: Drs. Fuzhong Weng, Banghua Yan, & Andrew Jones Purpose: - Improve atmospheric profiling & radiance assimilation over land - Improve the Land Surface Temperature (LST) - Reduce rejects over desert & elevated terrain Milestones: - Developed IR emissivity & single-channel LST retrieval - Implemented MEM into NAVDAS framework - Performed initial 1dvar emissivity retrieval -Combined infrared & microwave sensors into a single retrieval -Submitted working code to NRL beta-testing platform Future Plans: - Complete CRTM implementation - Transfer methodology to NCEP - Mesoscale impact studies (UAE 2 ) - Extend to SSMI/S assimilation - Complete strategy for bias correctionSummary