Mapping Climate Thresholds

The source data was obtained from ZedX, Inc. as part of their Hi-Rez Data – Climatological Series – and which consisted of daily maximum temperatures for the State of California for the period of January 1, 1992 through April 29, 2002 inclusive.  The temperature data (in degrees Fahrenheit) were interpolated by ZedX (using their proprietary algorithms) from point data into signed 16-bit raster data that were gridded to 1 sq. km. pixels to fit the boundary of the state.

The 3,772 data sets were imported into a GIS-ready format using the program ERDAS IMAGINE® version 8.6, a product of Leica Geosystems Geospatial Imaging.  As it was desired to determine which locations within California can experience temperature patterns which would be stressful to the Whitefly, subsets of the GIS-ready raster data were processed to generate specific products for preliminary analysis.

The specific products were derived using moving windows of three and five days and also, the total number of days, in which selected temperature criteria were met during the month of August, 2001.  The three temperature criteria examined were temperatures greater than or equal to 95, 100 and 105 ºF, respectively.

For example, the three-day moving window product was generated by examining rasters for three sequential days, Day 1, Day 2, and Day 3.  If a pixel was found to have met the temperature criteria for each of the three days, then a “one” was placed into the output raster at that location.  If the temperature criteria was not met for the three days, then a “zero” was placed into the output raster.  The output raster would therefore be a binary raster of those pixels which had met the temperature criteria.  The next sequence would be Day 2, Day 3, and Day 4, followed by Day 3, Day 4, and Day 5, each generating their own binary output raster.  After all of the binary output rasters were generated, they were summed to determine how many three-day periods met the specified temperature criteria during the month of August 2001.  This resulted in three output rasters, one for each of the three temperature criteria. This process was repeated for the five-day windows for each of the three temperature criteria and which generated three output rasters.  Lastly, the binary rasters corresponding to the three temperature criteria were simply summed to determine the total number of days during the month of August 2001 in which the specified temperature criteria occurred.  This resulted in three output rasters denoting the number of days in which the temperature criteria had been met.