INTEGRATION OF SHORT-DURATION GROUND BASED MEASUREMENTS WITH LONG-DURATION SATELLITE BASED MEASUREMENTS AND NUMERICAL WEATHER SIMULATIONS TO PROVIDE A MORE ACCURATE ASSESSMENT OF SOLAR ENERGY AVAILABILITY AND VARIABILITY

摘要

The energy output from a solar energy project depends largely on the strength of the solar resource at the site. Therefore, an accurate understanding of this resource is crucial for estimating project performance. This understanding has typically been important in financial calculations that consider the statistical properties of the annual and monthly average energy resource. However, a more detailed understanding is increasingly becoming important for designing energy storage or performing integration or interconnection studies of large solar facilities. All of these studies, from the most basic to the most advanced, benefit from having accurate and long-term time series of solar resource potential data.To provide resource estimates, ground based observations of global horizontal irradiance (GHI) and direct normal irradiance (DNI) are the most accurate. However, the number of monitoring stations available around the world is fairly small and sensor maintenance issues (e.g. cleaning of dust) make data collection difficult and costly. Satellite based estimates of shortwave radiation (both GHI and DNI) have recently been used as an alternative to ground-based sensors. These estimates are typically available for long periods of time (at least a decade) and over the majority of the planet at fairly high spatial resolution (approximately 3 km). However, what they make up for in spatial and temporal coverage, they lack in accuracy. Satellite based estimates are especially poor estimators of Direct Normal Irradiance.This paper presents a methodology that combines the accuracy of the well-maintained ground-based measurements and the temporal coverage of satellite-based methods to obtain more accurate long-term estimates of solar resource at a location. This method can effectively fill in the missing gaps of the ground based measurements (due to sensor maintenance issues) and also extend the ground-based record to provide a more complete picture of long-term variability of the resource at the site. The method alsoincorporates guidance from a numerical weather prediction (NWP) model that is also available globally and for the same period of record as the satellite measurements.We show results of this method applied to a single site in the United States (Desert Rock, Nevada) for the estimation of DNI and GHI. More information about the site can be obtained at http://www.srrb.noaa.gov/surfrad/desrock.html. For the purposes of this paper, we will use ground-based observations from the site during the year of 2008 to calibrate the model. DNI and GHI will then be predicted for 2006 and 2007 using the satellite and NWP estimates of site weather and irradiance.