7.9 The utility of a well-calibrated, high-resolution GOES-derived solar insolation and evapotranspiration database for water management over Florida, USA

Wednesday, 30 April 2008: 11:30 AM
Floral Ballroom Jasmine (Wyndham Orlando Resort)
John R. Mecikalski, University of Alabama, Huntsville, Huntsville, AL; and S. J. Paech, D. Sumner, and C. Pathak

A daily, solar insolation database has been developed over the period, 1995-2007 towards estimating evapotranspiration (ET) over the state of Florida at 2 x 2 km resolution. The primary component used in estimating incoming solar insolation is 30-min visible-channel data from Geostationary Operational Environmental Satellites (GOES)-8 and 12, along with radiative transfer modeling. The insolation estimates are combined with standard meteorological information via the Penman-Monteith and Priestley-Taylor equations to provide a high-quality spatially- and temporally-distributed dataset of reference and potential ET.

An extensive network of (pyranometer) surface measurements of solar radiation in Florida has allowed refined calibration of the GOES-derived daily-integrated insolation data product. This refinement of radiation data allowed for corrections of satellite sensor drift, satellite generational change and consideration of the highly variable cloudy conditions that are typical of State of Florida. In addition, detailed in situ to satellite comparisons were made using 47 field stations located throughout the State, along with simulated values of clear-sky solar radiation. This calibration occurred in three-steps: (1) comparisons with ground-based pyranometer measurements on clear (non-cloudy) reference days, (2) correcting for a low-bias that occurred under cloudy conditions, depicted by the ratio of an intermediate GOES product to clear sky solar radiation, and (3) deriving a monthly bias correction factor. Pre-calibration results indicated good model performance, with a station-averaged model error of 25 W m-2 day-1 (13%). Calibration reduced errors to 20 W m-2 day-1 (10%), and also removed time- and season-related biases. Data from an additional 9 pyranometers stations were used for a successful validation of the calibration strategy.

The final dataset will be used by State of Florida Water Management Districts to produce State-wide, 2 x 2-km resolution daily ET estimates that will aid in optimizing water resources management. The final portion of this presentation will outline these activities.

- Indicates paper has been withdrawn from meeting
- Indicates an Award Winner