School of Earth Sciences - Theses
Permanent URI for this collection
Search Results
1 - 1 of 1
-
ItemHigh resolution solar exposure estimates from geosynchronous satellite observations for climate and near real-time applications( 1998)This thesis describes research and development work undertaken to produce a satellite based near-real time high resolution (6 to 24 km) surface solar exposure estimation system. Physical models of radiative transfer within the atmosphere have been developed to produce the estimates of exposure for the entire Australian continent from full resolution hourly visible Geostationary Meteorological Satellite (GMS) Stretched-Visible and Infrared Spin Scan Radiometer (S-VISSR) data. This thesis describes the exposure estimation system, including details of the physical processes modelled. The accuracy of the exposure data is presented. The first high resolution climatology of exposure across Australia is also presented and discussed. Detailed charts of mean daily exposure for each month and annual mean daily exposure form part of the climatology, based on the period November 1990 to June 1994 inclusive. Annual and four-monthly charts are compared to the available Australian Bureau of Meteorology (BoM) National Climate Centre (NCC) Solar Radiation Atlas (1975) charts based on cloud and sunshine records for the period 1968 to 1974 inclusive. Generally the agreement is good, with the satellite system providing greater spatial and temporal (all months) detail, some significant differences and higher accuracy. The satellite climatology shows that minimum exposure in the far north occurs in February due to the monsoon even though the sub-solar point is at a similar latitude. During the monsoon, the exposure minimum over Cape York is seen to shift from the east to the west side. Considerable detail of coastal and orographic exposure gradients about the east and southeast coasts is available. Other features seen for the first time are also presented. For exposure estimation, the model of Diak and Gautier (1983) has been developed further and carefully tuned for use with GMS-4 data (1990 to 1994). Extensive changes have been made to this model to use data from GMS-5, which replaced GMS- 4 in May 1995. GMS-5 has a sensor response extending from the visible to the near-infrared. The GMS-5 based model now runs operationally within the BoM, using total precipitable water estimates from the BoM regional numerical weather prediction (NWP) system, and real-time ozone estimates from the local readout of the National Oceanographic and Atmospheric Administration (NOAA) satellites. The models perform best in clear-sky conditions, with the average deviation of spatially-averaged daily model estimates from surface-based point pyranometer data being less than 5% (less than 4% against available high quality pyranometer data). In cloudy conditions, the average percentage deviation is larger. Australia-wide estimates of the accuracy of satellite-based exposure estimates have been developed. Over most of the continent, typical cloud conditions lead to daily estimates being within 8% of collocated point pyranometer measurements. No other high-resolution data set is available for direct comparison. However, results achieved here are comparable to or better than those reported for other locations. In clear-sky conditions, results presented here are as accurate as measurements from well-maintained good-quality pyranometers. The spatial and temporal variability of the exposure data has also been examined. From this, it has been estimated that over typical Australian agricultural areas, daily satellite exposure estimates are more accurate than extrapolation from a high-accuracy pyranometer more than 20 to 50 km distant. The exposure data have already been used for crop modelling purposes, as an aid to siting of high-quality ground-based measurements for a solar-thermal power station feasibility study, and for hydrological modelling. Such applications, while briefly discussed, are outside the focus of this thesis.