343-1 Calibrating NASA POWER Data for Site-Specific Crop Modeling Using Minimum Observed Data.

See more from this Division: ASA Section: Climatology & Modeling
See more from this Session: Agroclimatology and Agronomic Modeling. II. Crop Growth Models and Instrumentation.
Wednesday, October 24, 2012: 8:15 AM
Duke Energy Convention Center, Room 264, Level 2
Share |

Justin Van Wart, Agronomy and Horticulture, University of Nebraska-Lincoln, Lincoln, NE and Kenneth Cassman, Agronomy and Horticulture, University of Nebraska, Lincoln, NE
Crop models can be used to establish long-term average trends in crop production, useful as yield goal benchmarks, predictors of current season yields, and in determining optimal fertilizer and irrigation applications. These crop models are typically driven by daily weather data, but weather data records of sufficient historical length and quality for simulating long-term averages are rarely available. Historical, daily, observed weather data are publically available for over 75,000 stations through the National Oceanic and Atmospheric Administration, though many of these stations only report a few years of data. At issue is how many years of observed data are required to derive plausible, complete, historical weather data records. Previous studies have used available data to generate entirely unobserved synthetic weather data, which maintain monthly mean and standard deviations of observed data. In this paper, we present a new method for “filling-in” un-reported weather data using the globally gridded NASA POWER weather database. Regression relationships established between NASA POWER and available station data are used to “fill-in” missing station weather data. This method is tested using 1, 2, 4 and 8 years of observed weather data to create 30 years of “filled-in” data at 5 stations across the US Corn Belt. Results are compared against 30 years of site-specific observed and generated synthetic weather data. Initial results indicate that as few as 2 years of observed data are required to “fill-in” long-term data, which outperform synthetic weather data as input in crop simulation models.
See more from this Division: ASA Section: Climatology & Modeling
See more from this Session: Agroclimatology and Agronomic Modeling. II. Crop Growth Models and Instrumentation.