Between 1951 and 1975, the National Weather Service had about 60 observatories measuring total horizontal irradiance (GHI). This group of observatories, although they have their own names, is often referred to as the SOLRAD network. Some observatories record continuously on recorder tape, while others only record the total daily energy (insolation) data received. Many of these data were digitized at the observatories and sent to regional centers (and then centrally to the National Climatic Data Center, or NCDC) (1995 edition of the NSRDB Database User Manual 1961-1990, 1995, according to operating instructions revised over the years), according to operating instructions revised over the years. “1991-2005 NSRDB Database (Updated Edition) User’s Manual” published in April 2007). While some quality control measures are also employed in these records, they contain a wide variety of systematic errors due to several instrumental and calibration problems. When these data were assessed, only 26 observatories provided data suitable for storage in national databases.
In the mid-1970s, people began to create a dataset for evaluating the performance of solar energy systems, later the NSRDB database. In 1977, the Solar Meteorology (i.e. SOLMET) database was created based on data from 26 SOLRAD observatories (which measured horizontal total irradiance data from 1951-1975), and from another 222 observatories (with extensive meteorological data, available The ERSATZ (Modeled or Synthesized) Solar Database was created with data for assessing solar energy. Most SOLMET/ERSATZ databases have data recording periods from July 1, 1952 to December 31, 1975.
SOL.MET combines all available solar radiation data and meteorological data and expresses them uniformly in International Standard Units (ie Sl units). The time of the data is in true solar time or local standard time, and the time of the weather observation is provided so that users can select the weather observation station closest to their chosen solar time or local standard time. The SOLMET data were adjusted to reflect the 1956 change in the International Direct Heliometer Scale (IPS) and 2% of the solar radiation measurements were added. This change brings the European irradiance meter into line with the American irradiance meter. Before 1976, calibration errors and other related errors in the GHI data used for the analysis of the SOLMET database had also been corrected using the clear sky-sun-noon (CSN) technique. Modeling the CSN values aims to create a set of standard annual irradiance (SYI) values. Calculated using long-term monthly precipitation data and turbidity data (SOLMET 1978; SOLMET 1979). Whenever a clear sky is observed at solar noon, the measured solar irradiance data can be compared with the modeled SYI values and the difference used to determine the overall calibration (correction) factor for the total pyranometer. Correction coefficients appearing between CSNs are obtained using linear interpolation. While this approach helps remove systematic calibration errors in the dataset, it also removes possible long-term trends in the dataset. Using the observed meteorological values, gaps or missing data were filled by the model. The models used to fill gaps or missing data in the SOLMET database can also be used to estimate solar irradiance at ERSATZ sites. Models for all irradiance values in the ERSATZ database are derived from meteorological observations. Within the SOLMET/ERSATZ database, DNI and DHI are primarily modeling data. Using GHI data from five observatories (Albuquerque, New Mexico, Fort Hood, Texas, Livermore, California, Maynard, Massachusetts, and Raleigh, North Carolina) and The DNI data establishes a regression equation to calculate the direct radiation value from the horizontal total radiation value (Randall and EWhitson Jr. December 1, 1977).