Data processing: measuring – calibrating – or testing – Measurement system in a specific environment – Earth science
Reexamination Certificate
2002-10-01
2004-06-08
McElheny, Jr., Donald (Department: 2857)
Data processing: measuring, calibrating, or testing
Measurement system in a specific environment
Earth science
Reexamination Certificate
active
06748327
ABSTRACT:
BACKGROUND OF THE INVENTION
1. Field of the Invention
The present invention relates to a method for estimating solar radiation flux density and other solar-correlated meteorological variables at a location, based on measurements of the same variables at other locations within the same geographic region for the same period of time. The description of the method is based on hourly measures of the variables but could be used with measurements representing lesser or greater length daytime periods. The method is similarly described using measurements taken 2 meters above the earth's surface but could be applied to measurements from other distances above the surface. Modeling or extrapolating variables to locations within the region, for which measurements are not available, is typically done using traditional statistical methods such as least squares regression of linear equations of a chosen order, or by employing the techniques of geostatistics, common to hydrology and geology. The later methods are characterized by providing preferential significance to measurements near the location to be estimated. Application of geostatistics to biometeorological applications has been limited due to poor correlations using common geophysical coordinate systems. The method of this invention improves the application of geostatistical techniques to biometeorological estimations through the application of three steps: (1) decomposition of any measurement into two components, the mean measurement from historical observations for the same period of the year and the normalized departure from the mean for the measurement, (2) decomposition of the measured solar radiation flux density into the product of the calculated extraterrestrial radiation at each location and the clearness index, (3) and, transformation of the coordinate system for measured and estimated mean value locations from (x,y,z), typically latitude, longitude, and elevation or depth, to a computed coordinate system (x′,y′,z′) based on the relative calculated extraterrestrial radiation within the region, and one or two additional regionally-specific axes for analysis and modeling of mean values. For a specific (x,y) location, the extraterrestrial radiation flux density is the same each year, on the same day-of-the-year and hour-of-the-day, i.e. a constant that can be calculated.
The value of this method would, at least, provide important information for the engineering of solar photovoltaic applications, agricultural and landscape irrigation management, regional water use allocations, and, passive space conditioning. A person knowledgeable in the fields of geostatistics and biometeorology should be able to implement this method and achieve results superior to other methods of spatial estimation of solar radiation flux density and other solar-correlated meteorological variables.
2. Description of the Prior Art
The method is illustrated for an example region using the synoptic observations made available by the California Irrigation Management Information System (hereafter CIMIS). This network of more than 180 stations depicted in
FIG. 1
of this application are scattered throughout the state and the associated database of measurements have been maintained and quality assured since 1982 by the State of California, Department of Water Resources. Each dot on the map of California county boundaries identified in
FIG. 1
represents the location of one CIMIS weather station. Details of the network, station locations, and the instruments at each station are described in
Technical Elements of CIMIS, The California Irrigation Management Information System
published December, 1998 by State of California, The Resources Agency, Department of Water Resources, Division of Planning and Local Assistance. Although the method is demonstrated for this network within the region identified as California, it could be applied to any region supporting a similar synoptic network of meteorological instruments and a database of historical observations. Similarly but not demonstrated, measurements need not be surface-based. Remotely sensed measurements, such as satellite-based images shown in
FIG. 8
, could be used in the application of this method. Spatial estimation of a variable is commonly done at grid intersection points on a uniform grid created to cover the entire region. From a high-density grid description, utility software can create a contour map of variable value ranges throughout the region, as in FIG.
2
.
FIG. 2
was created from a data file representing measurements of solar radiation flux density in Watts per square meter on a flat surface at 2 meters for Nov. 10, 1996 from 12 pm to 1 pm Pacific Standard Time. The traditional technique used was linear regression of a cubic equation to the dataset. The range of measurements varied from 89 to 725 Watts per square meter but this traditional method estimated negative values of solar radiation for some parts of the region. For sizing of photovoltaic applications, yearly average solar radiation flux density for every daylight hour of the year would be required.
FIG. 3
represents the same traditional technique used in
FIG. 1
, but applied to the average solar radiation flux density for November 10, 12 pm to 1 pm Pacific Standard Time for all years of record. The appearance of significant areas represented by negative solar radiation would question the application of the technique. A more sophisticated technique would be to consider applying an equation of a higher order or geostatistical techniques to the problem.
Geostatistical analysis is the collection of statistical and other numerical techniques for determining spatial correlations between measured variables within a region, developing models that represent those spatial correlations, and then using those models to estimate variables at other locations within the region. Geostatistics has been widely used in many fields to estimate variables at locations lacking measurement, but most commonly in hydrology and geology.
The first step in a geostatistical analysis would be an attempt to identify spatial correlation structures in the data between pairs of measurements with similar vector separations. Prior to the analysis of correlation however, the number of pairs of samples available for all possible similar separations must be computed and analyzed.
FIG. 4
indicates the number of pairs at various intervals of separation available in the database of observations for geostatistical analysis within the region. For a measure of correlation to be calculated for samples with a similar separation, there must be a sufficient number of pairs of samples. Note the value of 549 in the center of the map (FIG.
4
). This indicates that there are 549 sample pairs separated by less than ½ degree from approximately 150 stations in 18 years of samples for this hour-of-the-year. Consider as an example pair of stations, San Jose located at (−121.95, 37.326) and Sacramento located at (−121.218, 39.65). The vector separation from San Jose to Sacramento would be (0.732, 2.324) which would fall within the interval (1,2). In
FIG. 4
at that grid intersection, there are 65 sample pairs with a similar separation. Note also that at (−1,−2) there are also 65 samples. This sample pair would also be included in those 65 pairs and represents the vector drawn from Sacramento to San Jose, which is equal in magnitude but in the opposite direction of the San Jose to Sacramento vector. Geostatistics accommodates this double counting of sample pairs. As indicated in
FIG. 4
, it is very important to note that there are only an adequate number of samples for some separations only out to about 3.5 degrees. Hence, further analysis of spatial correlations for these measurements will be limited to separations of 3.5 degrees or less for any direction. Alternatively, this would represent approximately 35% of the maximum separations possible within the region.
The correlation between sample pairs for all possible spacings within t
Lampe Thomas R.
McElheny Jr. Donald
LandOfFree
Method for spatial estimation of solar radiation flux... does not yet have a rating. At this time, there are no reviews or comments for this patent.
If you have personal experience with Method for spatial estimation of solar radiation flux..., we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and Method for spatial estimation of solar radiation flux... will most certainly appreciate the feedback.
Profile ID: LFUS-PAI-O-3318301