Comparison of NASA AIRS and MODIS Land Surface Temperature and Infrared Emissivity Measurements from the EOS AQUA platform Robert Knuteson, Steve Ackerman, Hank Revercomb, Dave Tobin University of Wisconsin-Madison CIMSS 08 April 2008
Imager and Sounder Pairs (visible, infrared, microwave) POES AVHRR HIRS/AMSU (1977 2015?) EOS MODIS AIRS/AMSU (2002 2010?) METOP AVHRR IASI/AMSU (2006 2020) NPP VIIRS CrIS/ATMS (2009 2016) NPOESS VIIRS CrIS/ATMS (2015 2025+) GOES Imager - Sounder (1975 2018?) GOES-R ABI - ABS(?) (2018? 2025+) (Others: EUMETSAT, Japan, China, India, ) - Denotes Operational Sensors (24/7 continuous operation)
Imager/Sounder Synergy Imager Contiguous spatial coverage. IR Sounder - Vertical profile at lower spatial res MW Sounder - All weather at lower spatial res. Ideally a single algorithm would combine these data optimally. But at a minimum, comparison of products from imager and sounder data should be used as a quality check on potential CDRs (climate data records). This presentation is a preliminary imager/sounder assessment using EOS MODIS and AIRS products as a prelude to the operational sensors on NPP/NPOESS.
Is this a Climate Comparison? No. Aqua AIRS and MODIS record is less than five years. Instead the goal of this effort is to evaluate the following: What are the natural spatial and temporal scales of the natural variability of the relevant quantities? To what degree can we identify BIASES in the LST product? When product algorithm changes are made (i.e. version changes), do we have a way of deciding if the intended improvements actually improve or degrade the product accuracy? These are elements of data stewardship and so appropriate for consideration at this NCDC workshop.
Why is Land Surface Temperature (LST) important to sounding over land? (Importance to the AIRS science team)
AIRS FOV 15 km Atmospheric IR Sounder (AIRS)
AIRS EOS AQUA Water Vapor PWV Day This is monthly average but AIRS provides nearly complete daily coverage.
AIRS PWV Night
AIRS PWV Day - Night Are Day/Night Total Water Vapor Retrievals impacted by surface? (yes)
AIRS Cloud Clearing Method Cloud Fraction Day
AIRS Cloud Fraction Night
AIRS Cloud Fraction Day - Night Cloud Clearing increases coverage (from global monthly to daily) but how well is this method working over variable surfaces!?!
AIRS Tskin Day Same multi-spectral AIRS LST algorithm used over ocean and land.
AIRS Tskin Night
AIRS Tskin Day - Night Land Day/Night Ts Variations are > 40 K in sparsely vegetated areas!!!
Comparison of AIRS and MODIS land products from the NASA EOS Aqua Platform
MODIS and AIRS LST Climate Product Characteristics Potential Sources of Bias and Mitigation Approaches Aqua MODIS Aqua AIRS Sensor Calibration < 0.2 K (windows) < 0.2 K Atmospheric Attenuation Column Retrieved Profile Retrieved Cloud Contamination Cloud Detection Cloud Clearing Surface Type Multi-spectral (004) Multi-spectral Land Cover Class (005) Temporal Sampling Clear only; 1:30 AM, PM Partly Cloudy; 1:30 AM, PM and Resolution (0.333 msec per sample) (30 msec per sample) Spatial Sampling 1 km Clear Only Samples 45 km CC (< 60% CF) and Resolution (1 km 5 km > 1 deg) (15 km 45 km > 1 deg)
Uniform Collocated MODIS & AIRS Band 34 13.7 micron Tobin, et al 2006
AIRS- MODIS Brightness Temperature Comparison (Tobin et al, 2006) Window Channels < 0.2 K Tobin, D. C., H. E. Revercomb, C. C. Moeller, and T. S. Pagano (2006), Use of Atmospheric Infrared Sounder high spectral resolution spectra to assess the calibration of Moderate resolution Imaging Spectroradiometer on EOS Aqua, J. Geophys. Res., 111, D09S05,
MYD11C3 Aqua MODIS (Early PM orbit) Monthly Average Clear Day/Night 0.05 degree grid (5 km x 5 km) QC Flag = 00 Collection 004 Sep 2002 Dec 2006 Collection 005 Sep 2002 Sep 2005 & Jan 2007 Nov 2007 AIRS and MODIS Collection 004 Collection 005 15-Sep-2002 15-Sep-2002 15-Nov-2002 15-Oct-2002 15-Dec-2002 15-Nov-2002 15-Jan-2003 15-Dec-2002 15-Feb-2003 15-Jan-2003 15-Mar-2003 15-Feb-2003 15-Apr-2003 15-Mar-2003 15-May-2003 15-Apr-2003 15-Jun-2003 15-May-2003 15-Jul-2003 15-Jun-2003 15-Aug-2003 15-Jul-2003 15-Sep-2003 15-Aug-2003 15-Oct-2003 15-Sep-2003 15-Dec-2003 15-Oct-2003 15-Feb-2004 15-Dec-2003 15-Mar-2004 15-Jan-2004 15-Apr-2004 15-Feb-2004 15-May-2004 15-Mar-2004 15-Jun-2004 15-Apr-2004 15-Jul-2004 15-May-2004 15-Aug-2004 15-Jun-2004 15-Sep-2004 15-Jul-2004 15-Oct-2004 15-Aug-2004 15-Nov-2004 15-Sep-2004 15-Dec-2004 15-Oct-2004 15-Jan-2005 15-Nov-2004 15-Feb-2005 15-Dec-2004 15-Mar-2005 15-Jan-2005 15-Apr-2005 15-Feb-2005 15-May-2005 15-Mar-2005 15-Jun-2005 15-Apr-2005 15-Jul-2005 15-May-2005 15-Aug-2005 15-Jun-2005 15-Sep-2005 15-Jul-2005 15-Oct-2005 15-Aug-2005 15-Nov-2005 15-Sep-2005 15-Dec-2005 15-Jan-2007 15-Jan-2006 15-Feb-2007 15-Feb-2006 15-Mar-2007 15-Mar-2006 15-Apr-2007 15-Apr-2006 15-May-2007 15-May-2006 15-Jun-2007 15-Jun-2006 15-Jul-2007 15-Jul-2006 15-Aug-2007 15-Aug-2006 15-Sep-2007 15-Nov-2006 15-Oct-2007 15-Dec-2006 AIRS / MODIS Matched Months Month Jan Feb Mar Apr May June July Aug Sept Oct Nov Dec 004 Total 47 47 15-Nov-2007 3 4 4 4 4 4 4 4 4 3 4 5 005 4 4 4 4 4 4 4 4 5 4 3 3
MODIS NIGHT Require > 50% LST coverage of 1 degree grid cell with QC = 00.
Std Deviation 1 deg LST MODIS NIGHT Spatial variability only. No variation over month is retained! (***)
Mean 1 deg LST MODIS NIGHT
MODIS AIRS NIGHT Barren land shows MODIS cold bias (collection 005) up to 8 degrees.
MODIS DAY Require > 50% LST coverage of 1 degree grid cell with QC = 00.
MODIS DAY Spatial variability only. No variation over month is retained! (***)
MODIS DAY
MODIS AIRS DAY Barren land shows MODIS cold bias (collection 005) up to 10 degrees.
Use Land Classes (IGBP) to group the global data by land type for statistical analysis. IGBP CLASS ID 0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 IGBP CLASS Description Water Bodies Evergreen Needleleaf Forest Evergreen Broadleaf Forest Deciduous Needleleaf Forest Deciduous Broadleaf Forest Mixed Forest Closed Shrublands Open Shrublands Woody Savannas Savannas Grasslands Permanent Wetlands Croplands Urban and Built-Up Cropland/Natural Vegetation Mosaic Snow and Ice Barren or Sparsely Vegetated Missing Data
Forests NIGHT Barren/Savanna MODIS 004 Grassland/cropland Ice/Snow
Forests NIGHT Barren/Savanna MODIS 005 Grassland/cropland Ice/Snow Note artifacts ( noise ) is reduced in monthly timeseries.
MODIS 004 NIGHT Snow/Ice Covered Land Warm clouds over cold snow/ice contaminate the AIRS LST monthly product.
MODIS 004 NIGHT AIRS and MODIS (collection 004) agree to within 0.5 K at night!!!
MODIS 005 NIGHT MODIS (collection 005) is 0.5 2.5 K colder than collection 004?
AIRS and MODIS (collection 004) agree to between 0 and -1.5 K in the Day.
MODIS (collection 005) is 0.5 2.5 K colder than collection 004?
MODIS Collection 004 (red) Note: AIRS day/night emissivity bias will be addressed In next version update. 11 μm B31
MODIS Collection 005 (red) Collection 005 emissivity is assigned to a higher value than was determined in Collection 004. 11 μm B31
MODIS 004 NIGHT AIRS 12 μm B32 11 μm B31 Single Month MODIS MODIS AIRS 8.5 μm B29 3.9 μm B20 MODIS MODIS
MODIS 005 NIGHT AIRS 12 μm B32 11 μm B31 Single Month MODIS MODIS AIRS 8.5 μm B29 3.9 μm B20 MODIS MODIS
Summary: AIRS vs MODIS LST MODIS collection 004 Clear Day/Night algorithm and AIRS (version 5) cloud-cleared multi-channel regression retrieval agree to within 0.5 K at night (!!!) [excluding snow/ice covered land]. Between 0 and -1.5 K during the Day. MODIS collection 005 Clear Land Classification algorithm is 0.5 to 3 degrees colder than collection 004 (and colder than AIRS v5). The change in assumed MODIS emissivity is consistent with this LST change. This raises a question about land cover classification schemes and how to avoid building in systematic biases. The fact that biases can be assessed through a comparison of AIRS and MODIS suggests that a continuous comparison of imager and sounder LST products will be a useful quality check on future operational algorithms.