Method for spatial estimation of solar radiation flux density and other solar correlated meteorological variables

A multivariable statistical method for use when making decisions relating to irrigation management, photovoltaic application engineering, passive space conditioning or other activities, for estimating solar correlated variables for a geographical location identified by latitude and longitude within a geographical region including a synchronous network of a plurality of weather stations incorporating instruments necessary to acquire or calculate the variables and a database of historical meteorological values of the variables for specific time periods over a number of years, said method comprising the steps of:

• selecting from the database of said plurality of weather stations a dataset including data for the variables for a time period identified by day of year, time of mid-period, and period length for the same specific time period of every calendar year for a number of years;
• calculating for every occurrence of measured solar radiation in the dataset the clearness index and the extraterrestrial radiation as a function of latitude, longitude, time zone, day of year, time of mid-period, and period length;
• calculating for every weather station in the dataset identified by latitude and longitude a primary coordinate in an alternative coordinate system whose value is the relative extraterrestrial radiation for the time period as a fraction of the maximum extraterrestrial radiation at any location within the geographical region for the time period;
• calculating for every weather station in the dataset identified by latitude and longitude one or more coordinates in an alternative coordinate system whose value is a relative distance from the weather station to a boundary specific to the geographical region as a fraction of the maximum distance from any location within the geographical region to the boundary;
• calculating the mean value for each variable for each weather station identified by latitude and longitude in the dataset;
• calculating the departure from the mean for every recorded variable value for each weather station identified by latitude and longitude in the dataset;
• calculating the normalized departure from the mean by dividing every departure by the mean by the weather station mean value for the same recorded variable;
• analyzing spatial correlations and developing variogram models for the mean variables in the alternative coordinate system;
• analyzing spatial correlations and developing variogram models for normalized departure variables in latitude and longitude;
• creating a data file containing the latitude, longitude, and equivalent coordinates in the alternative coordinate system that defines grid intersections on a uniform mathematical grid over the region with selected resolution;
• for a preselected variable, estimating the mean for every grid intersection identified by the equivalent coordinates in the alternative coordinate system contained in the data file by kriging or other appropriate geostatistical modeling method utilizing the variogram models developed;
• for said preselected variable, estimating the normalized clearness departure for every grid intersection identified by latitude and longitude contained in the data file by kriging or other appropriate geostatistical modeling method utilizing the variogram models developed;
• fitting a linear model defined by slope and intercept to the relationship between all calculated normalized departures for the preselected variable and all normalized clearness departures for every station in the dataset;
• calculating the estimate of the normalized variable departure at every grid intersection defined in the data file utilizing the linear model applied to the normalized clearness departure for the same grid intersection;
• and
• at every grid intersection summing the estimate for the mean and the estimate for the departure from the mean for the preselected variable to obtain the variable estimate at each grid intersection.