Methods
We downloaded 5 years (2000-2004) of daily surface data from the NCDC online climate data clearinghouse. The downloaded data included dozens of climate variables recorded for more than 19,000 stations nationwide. We processed the data to extract four variables: total precipitation, minimum and maximum temperature, and relative humidity. For each station, we tallied (1) the total number of days and (2) the longest number of consecutive days in a given year that met the following conditions: maximum temperature greater than 60 °F, minimum temperature less than 80 °F, and at least a trace amount of precipitation or relative humidity of greater than 85 percent. These threshold values were selected to reflect current knowledge about the climatic conditions favorable for P. ramorum survival and spread.
We recorded the latitude, longitude, and elevation values for each weather station from an associated data set. We dropped any stations that fell outside the conterminous United States and any stations with more than 30 days of missing data for any variable in a given year. This filtering process reduced the number of usable stations (Table: Number of NCDC), but still yielded consistent national coverage. For stations missing 1-30 days of data, we normalized the total-day and consecutive-day count values by dividing them by the proportion of days in the year for which data were available and then rounding to the closest integer.
- Gradient Plus Inverse Distance Squared Interpolation : We interpolated gridded maps of the conterminous United States for both the total-day and consecutive-day variables using a gradient plus inverse distance squared (GIDS) approach.
- Evaluation : For comparison to the GIDS-derived total-day and consecutive-day count maps, we created gridded maps for 2000-2004 using two spatial interpolation methods available through the ArcGIS Geostatistical Analyst extension.
Encyclopedia ID: p3407