Modeling Current Climate Conditions for Forest Pest Risk Assessment

Forest pest risk assessments detail the nature and severity of threats posed to particular forest species and ecosystems by insects, pathogens, or other organisms (Andersen and others 2004a). With respect to nonindigenous forest pests, risk can be categorized or quantified based on a combination of factors: the potential for the pest to become established, the potential for it to spread following introduction, the potential to cause economic damage, or the potential to cause environmental harm (NAFC 2004). A commonly desired product of such assessments is a map depicting the threat posed by introduction or establishment of a forest pest throughout a geographic area of interest (Andersen and others 2004a). These maps can facilitate early detection and response procedures, providing a template for the design of regulatory programs and detection surveys. If a pest has already been established in one part of the geographic area of interest, threat assessment maps are used to help set control priorities for other geographic areas that are at high risk of invasion (Andersen and others 2004b).

Given the limitations of existing climatic data sets, we explored the use of NCDC daily weather station data for the United States as an alternate source for maps relevant to forest pest risk assessments. We had three basic objectives: (1) to spatially interpolate annual counts of the number of days with co-occurrence of multiple climatic variables relevant to the growth and spread of a specific forest pest—the pathogen that causes sudden oak death (Phytophthora ramorum); (2) to identify a spatial interpolation method appropriate for count-based data and compare it to some common geostatistical approaches; and (3) to assess the utility of the derived maps for depicting risk.