Observing Changes of Surface Solar Irradiance in Oregon: A Comparison of Satellite and Ground-Based Long-Term Time-Series

Significant increases over time are found in direct normal irradiance (DNI) in Oregon using both ground and satellite-derived measurements of DNI. Linear regression of all locations in both data sets shows strong positive trends of .4% to .6% per year. Ground measurements are analyzed from 1980 (and at one site from 1978) until 2004. These 25 years of ground measurements come from three climatically diverse sites in the state of Oregon using an Eppley Normal Incidence Pyrheliometer (NIP). The NIP is a good candidate for long term trend analysis as its responsivity remains consistent over time. The sensitivity of the Eppley Precision Spectral Pyranometer (PSP) which measures total radiation, on the other hand, degrades over time, approximately .5% to 2% per year. This uniquely long data set is compared to DNI calculated from the International Satellite Cloud Climatology Project (ISCCP). The ISCCP D series applied here has 280 km x 280 km boxes, each of which includes one of the ground based sites, giving cloud and atmospheric input data from 1983 until 2001. Radiative transfer calculations are done using the two-stream method from the library for radiative transfer (libRadtran). The three hourly satellite observations allow comparison of different time integration periods. Besides annual average comparisons, monthly averages are examined to look for seasonal variation and confirm that the observations show a regional trend. Ground measurements of DNI for this length of time are rare, making this study a unique opportunity to test the capability to calculate direct normal irradiance based on ISCCP results. The agreement of the ISCCP derived irradiances to the measurements is very good: the trends differ between .08 and .3 W/m{2 depending on the site. From 1998 through 2002 satellite data were used to produce a solar radiation database on a 0.1i° grid. Comparisons between the modeled beam irradiance for the coordinates of the ground based station will be compared to the average for the area of the ISCCP grid to check how representative each ground site is of the ISCCP box. The successful verification of ISCCP for this application at three independent sites in this region allows us to use this approach to also analyze similar changes over other regions. Comparing these two methods of obtaining direct irradiance also provides valuable information about the sources of seasonal and inter-annual changes in cloud cover and other atmospheric constituents.