University of Montana ScholarWorks at University of Montana Numerical Terradynamic Simulation Group Publications Numerical Terradynamic Simulation Group 2012 Validating a Satellite Microwave Remote Sensing Based Global Record of Daily Landscape Freeze- Thaw Dynamics John S. Kimball University of Montana - Missoula Youngwook Kim Kyle C. McDonald City College of New York Let us know how access to this document benefits you. Follow this and additional works at: https://scholarworks.umt.edu/ntsg_pubs Recommended Citation Kimball, John S.; Kim, Youngwook; and McDonald, Kyle C., "Validating a Satellite Microwave Remote Sensing Based Global Record of Daily Landscape Freeze-Thaw Dynamics" (2012). Numerical Terradynamic Simulation Group Publications. 369. https://scholarworks.umt.edu/ntsg_pubs/369 This Presentation is brought to you for free and open access by the Numerical Terradynamic Simulation Group at ScholarWorks at University of Montana. It has been accepted for inclusion in Numerical Terradynamic Simulation Group Publications by an authorized administrator of ScholarWorks at University of Montana. For more information, please contact scholarworks@mso.umt.edu.
Validating a Satellite Microwave Remote Sensing Based Global Record of Daily Landscape Freeze-Thaw Dynamics John S. Kimball 1, Youngwook Kim 1, Kyle C. McDonald 2,3 1 Flathead Lake Biological Station, Division of Biological Sciences, The University of Montana, USA. 2 Jet Propulsion Laboratory, CalTech, Pasadena CA. 3 The City College of New York, City University of new York AGU Fall Meeting (IN52B-01), December 7, 2012
The Freeze Thaw Earth System Data Record (FT ESDR) Objectives: Build a consistent global record of daily landscape freeze thaw (FT) status where frozen temperatures constrain ecosystem processes; Resolve FT heterogeneity in accordance with mesoscale climate & LC variability; Link FT processes with ecosystem productivity & C exchange; Distinguish FT seasonal/annual variability from longer term climate trends. Mean Non-frozen Season (days) Non-frozen Season Variation (SD, days yr -1 ) Global FT Seasonality Non-frozen Transitional Frozen Annual Variation Source: Kim et al. 2011. IEEE TGARS 49(3). Approach: Integration of satellite microwave 37V GHz T b records from SMMR & SSM/I; Temporal change classification of daily (AM & PM overpass) T b series using seasonal FT reference states on a grid cell wise basis; 4 discrete classification levels: F (AM & PM), NF, TR (AM frozen, PM non frozen) & INV TR. Domain: Global vegetated land areas where fozen temperatures constrain annual productivity. Status: 2010: Initial (V.1) data release (20 yr record); 2012: V.2 data release (32 yr record); FT ESDR publicly available (NSIDC DAAC); Additional updates & data releases planned. Documentation: Kim et al. 2011. IEEE TGRS 49(3) Kim et al. 2012 Rem Sens Environ 121 Funding: NASA MEaSUREs Program
Verification of FT ESDR Quality (QC) Daily (AM, PM) FT comparisons against surface air temperatures (T mn, T mx ) from ~3,700 WMO weather stations. Results aggregated to daily global mean spatial classification accuracy (%) & included with product QC metadata. Mean Annual Accuracy for Validation Sites Accuracy (%) Global Mean Annual Classification Accuracy Global Mean Daily Classification Accuracy Spatial SD
FT ESDR Quality Assessment (QA) QA Elements: Static multivariate empirical prediction of WMO station mean annual FT accuracy: Accounts for terrain, open water & landcover heterogeneity, & FT reference state uncertainty; Dynamic flagging of RFI, active precipitation & data gaps; Distinguishes dry soil climate areas where alternative FT algorithm is used; Rescaling to dimensionless (0 1) QA spatial distributions; QA distributions partitioned into discrete quality categories ranging from Poor (spatial classification accuracy < 70%) to Best (>90%) accuracy; QA records computed annually & provided with FT ESDR metadata. 2010 FT-ESDR QA (SSM/I) 2010 FT-ESDR QA (AMSR-E)
FT ESDR Validation: Cross Sensor Comparisons Document FT accuracy against similar retrievals from other satellite microwave records & relative to in situ station observations. Mean Annual Non-Frozen Season (2003-10) AMSR-E Mean Annual FT Classification Accuracy SSM/I Days WMO Validation Sites
FT ESDR Uncertainty: FT Sensitivity & Sub grid Heterogeneity Elements: FT sensitivity studies in relation to individual landscape elements, varying sensor frequencies & polarizations; Evaluate surface air, soil & vegetation components affecting the aggregate landscape FT signal; Clarify sub grid scale heterogeneity effects using overlapping finer scale sensor records. C 20 10 0 10 Aniak AK (2010) 20 30 AirT SoilT 40 0 30 60 90 120 150 180 210 240 270 300 330 360 Tb (kelvin) 10 0 10 20 30 40 50 60 70 0 30 60 90 120 150 180 210 240 270 300 330 360 SSM/I 37V (PM) SMOS (PM) AMSR E 36V (PM) ASCAT (PM) 5 6 7 8 9 10 11 12 13 14 15 ASCAT db (σ)
Verifying FT ESDR Variability & Trends Document FT variability & trends relative to synergistic information from global model reanalysis data. Northern Hemisphere Mean Annual Non- Frozen Season Trend (1979-2010) FT-ESDR Northern Hemisphere Non-frozen Season Trend (1979-2010) Trend: 1.9 d decade -1 (p<0.001) Days decade -1 >7.5 Mean annual T av (NCEP, NCEP2, MERRA) Annual Non frozen period (FT ESDR) T av uncertainty range <-7.5 Source: Kim et al. 2012. Rem. Sens. Environ. 121.
Evaluating FT ESDR Linkages to Ecosystem Processes The FT ESDR provides a surrogate measure of frozen temperature constraints to plant growth & the potential growing season, defined by satellite (MODIS, AVHRR) NDVI & tower measures of ecosystem productivity. Satellite FT & NDVI grid cell extractions over a selected (Boreal ENF) FLUXNET Tower site FT legend: 0=Frozen; 1=Non-frozen; 2=Transitional Source: Kim et al. 2012. Rem. Sens. Environ. 121.
Vegetation Response to FT ESDR Frozen Season Changes Positive relations between FT ESDR non frozen season variability & satellite (MODIS) based productivity metrics for energy constrained regions; FT effects reduced or reversed for moisture constrained areas; FT ESDR shows general relaxation of frozen temperature constraints to productivity consistent with other satellite & observational data records showing earlier/longer growing season trends. Correlation (r) between FT ESDR non frozen season & MODIS summer (JJA) NDVI growth anomalies (2000 2008) Temperature & Water constraints to NPP (GMAO MERRA) Temp Water
FT ESDR Status Recent data releases include: 32 year (SMMR & SSM/I) FT record (1979 2010); 9 year AMSR E FT & global land parameter bundle (2002 2011); Available online through FT ESDR project page (http://freezethaw.ntsg.umt.edu) & NSIDC DAAC; Additional data releases planned, including longer records, regional products & exploiting new observations & lessons learned from continuing validation efforts. Product characteristics: Global domain; continuous daily (AM, PM, CO) record; Formats: HDF 5, binary, Geotiff; detailed metadata documenting product quality & uncertainty; Product quality: Good; Validated Stage III (CEOS LPV guidelines): Mean annual FT classification accuracy >80% relative to global station observations; Established methods & documentation (Kim et al. 2011, TGARS 49; 2012 RSE 121); Accuracy & uncertainty verified against multiple sensor records & synergistic biophysical data; Additional validation activities ongoing.
Thank You! Collaborators: : John S. Kimball, Youngwook Kim, Joseph Glassy, Lucas Jones, Jinyang Du (UMT( UMT); Kyle McDonald, Ronny Schroeder, Marzieh Azarderakhsh (CUNY); Erika Podest,, Scott Dunbar, Bruce Chapman (JPL( JPL CalTech)