Raster Tools

Geography is central to forestry and natural resource planning because it shapes the relationships among forests, vegetation, soils, and landscapes. As a result, geospatial analytics play an important role in decisions about natural resources. To support that work, scientists developed Raster Tools, a free, open-source geospatial platform capable of processing large datasets and complex workflows. Its purpose is to efficiently integrate geospatial processing with the latest in parallel processing, remote sensing, machine learning, wildfire planning, and natural resources management.

In land management, scientists, analysts, and managers rely on data to characterize and study forests and rangelands. These data come in many forms, including tables, vector data, rasters and complex multidimensional datasets. Many of these data sources are extremely large, which makes them difficult to efficiently process and use in geospatial analyses. To convert this raw data into practical outputs such as maps, visualizations, and reports, specialized software is required. This software needs to handle traditional relational databases and more complex, nonstandard formats in a way that allows them to work together seamlessly on a personal computer and or cloud-based systems.

Photo Credit

Forest Service image by John Hogland

To overcome these technical obstacles, Forest Service and University of Montana researchers created Raster Tools, a free, open-source Python software package that uses significantly less memory, storage space, and processing time to perform geospatial and machine learning with large geospatial datasets. Raster Tools provides an application programming interface (API) to underlying code that uses delayed and automated parallel processing to facilitate a wide range of spatial, statistical, and machine learning analyses at scale, across both local and cloud computing platforms.

Raster Tools is accessed and installed using Python’s (pip) or Conda’s package manger and Python’s (PyPI) or conda-forge package indices. While Raster Tools can be integrated with other coding libraries to create a graphical user interface, it’s not a point-and-click dropdown tool. Instead, script builders and developers use Raster Tools’ API to analyze data quickly and efficiently. It works on any processing platform (Windows, Mac, and Unix) and can be used to answer complex natural resource related questions.

Raster Tools Applications

Forest Service research forester John Hogland and colleagues have developed various Raster Tools products with tutorials so analysts, GIS specialists, and land managers can learn how to use them to address pressing questions and tasks on their local units. The following analysis tools, each presented in the form of a Jupyter notebook, are examples of how Raster Tools can be used in natural resource management:

• Delivered Cost quantifies timber values based on harvest and transportation distances to a mill. It has also been used to map and analyze wildfire risk to landscapes and communities, coordinate wildfire response resources, and reduce burnable fuels.
• Fire Resilience models detailed fuels treatments at high specificity and shows the cost implications of specific wildfire mitigation treatments at operational spatial scales.
• Sample Design aids researchers and practitioners alike in designing well-spread and balanced field samples.
• Habitat Restoration helps identify and prioritize gopher tortoise restoration sites.

Finally, Raster Tools can be used to build other tools for a wide range of natural resources questions and applications.

Personnel experienced in working with vector and raster data within a geographic information system (GIS) can also access Raster Tools functionality through a graphical user interface using the recently developed QGIS plugin.

Purpose

Raster Tools is an easy-to-use Python software package that facilitates a wide range of spatial, statistical, and machine learning analyses that use large spatial datasets. Raster Tools can significantly reduce the processing time and storage space that is typically required. Researchers have developed an array of products with tutorials to help scientists, managers, and analysts get started using Raster Tools.

Applicability and Overview

Key uses: There are many potential applications, including fire; ecosystems; recreation; forest health, products, and management; and inventory, monitoring, and assessment

Scale: Local, mid-scale, and landscape