The Invasive Species Forecasting System

The Invasive Species Forecasting System (ISFS) provides computational support for the generic work processes found in many regional-scale ecosystem modeling applications. Decision support tools built using ISFS allow a user to load point occurrence field sample data for a plant species of interest and quickly generate habitat suitability maps for geographic regions of management concern, such as a national park, monument, forest, or refuge. This type of decision product helps resource managers plan invasive species protection, monitoring, and control strategies for the lands they manage. Until now, scientists and resource managers have lacked the data-assembly and computing capabilities to produce these maps quickly and cost efficiently. ISFS focuses on regional-scale habitat suitability modeling for invasive terrestrial plants. ISFS s component architecture emphasizes simplicity and adaptability. Its core services can be easily adapted to produce model-based decision support tools tailored to particular parks, monuments, forests, refuges, and related management units. ISFS can be used to build standalone run-time tools that require no connection to the Internet, as well as fully Internet-based decision support applications. ISFS provides the core data structures, operating system interfaces, network interfaces, and inter-component constraints comprising the canonical workflow for habitat suitability modeling. The predictors, analysis methods, and geographic extents involved in any particular model run are elements of the user space and arbitrarily configurable by the user. ISFS provides small, lightweight, readily hardened core components of general utility. These components can be adapted to unanticipated uses, are tailorable, and require at most a loosely coupled, nonproprietary connection to the Web. Users can invoke capabilities from a command line; programmers can integrate ISFS's core components into more complex systems and services. Taken together, these features enable a degree of decentralization and distributed ownership that have helped other types of scientific information services succeed in recent years.