Location: Hydrology and Remote Sensing LaboratoryTitle: Evaluation of SMAP freeze/thaw retrieval accuracy at core validation sites in the contiguous United States
|KRAATZ, S. - University Of New Hampshire|
|JACOBS, J. - University Of New Hampshire|
|SCHROEDER, R. - University Of New Hampshire|
|CHO, E. - University Of New Hampshire|
Submitted to: Remote Sensing
Publication Type: Peer Reviewed Journal
Publication Acceptance Date: 10/1/2019
Publication Date: 10/1/2019
Citation: Kraatz, S., Jacobs, J., Schroeder, R., Cho, E., Cosh, M.H., Seyfried, M.S., Prueger, J.H., Livingston, S.J. 2019. Evaluation of SMAP freeze/thaw retrieval accuracy at core validation sites in the contiguous United States. Remote Sensing. 10(9):1483. https://doi.org/10.3390/rs10091483.
Interpretive Summary: Freeze-thaw dynamics at the soil surface impacts many aspects of the energy and hydrologic cycles. Usually research studies focus on freeze-thaw systems that are in the boreal or very northern regions of the globe. However, freeze-thaw may play an extensive role in the soils within the contiguous United States, as well. A study was conducted at multiple watershed networks operated by the USDA throughout the U.S. and compared with the freeze-thaw products available from the Soil Moisture Active Passive (SMAP) mission. It was shown from remote sensing as well as modeling products available from NASA that freeze-thaw regularly extends to latitudes as far south as 35 degrees latitude and future studies should include this extended domain when considering regions of interest.
Technical Abstract: Seasonal freeze-thaw (FT) impacts much of the northern hemisphere and is an important control on the water, energy and carbon cycle. However, although FT in natural environments extends south of 45oN, FT studies using L-band have so far been restricted to boreal or greater latitudes. This study addresses this gap by applying a seasonal threshold algorithm to Soil Moisture Active Passive (SMAP) data (L3_SM_P) to obtain a FT product south of 45oN (‘SMAP FT’), which is then evaluated at SMAP core validation sites (CVS) located in the contiguous United States (CONUS). SMAP landscape FT retrievals are usually in good agreement with 0-5 cm soil temperature at SMAP grids containing CVS stations (>70%). The accuracy could be further improved by taking into account specific overpass time (PM), the grid specific seasonal scaling factor, data aggregation method and sampling error. Annual SMAP FT extent maps compared to modeled soil temperatures derived from the Goddard Earth Observing System Model Version 5 (GEOS-5) show that seasonal FT in CONUS extends to latitudes of about 35-40oN, and that FT varies substantially in space and by year. In general, spatial and temporal trends between SMAP and modeled FT were similar.