Project Proposal

The past few years have witnessed an upsurge of interest in climate variability, climate change, and global warming – themes that are addressed in several undergraduate and graduate courses at Oregon State University. Both faculty and students are interested in classroom-accessible data and tools that allow them to explore questions about climate variability. We are fortunate at OSU to have the PRISM (Parameter-elevation Regressions on Independent Slopes Model) group, which produces high-quality spatial climate data sets used by educators and students in universities and research groups across the country to study climate patterns. The PRISM group has made initial strides in developing web-based tools for exploration of climate data. However, questions about climate variability are becoming increasingly complex, and the need to perform advanced queries and analyses, such as those that involve moving freely among multiple spatial and temporal scales to discover patterns and extremes, often exceed the capabilities of existing tools. New tools are needed to help educators and students explore climate data and analyze climate variability.

Another problem is that fairly sophisticated levels of GIS (geographic information systems) and mapping expertise are typically required in order to access, analyze, and interpret the considerable variability in climate behavior at local to regional scales. Educators and students examining questions in areas such as land management, water scarcity, fire management, agriculture, forestry, tourism, and utilities may or may not have this level of expertise. Consequently, they request analysis tools that are not only increasingly sophisticated, but also easy-to-use, do not require desktop GIS software, and are capable of helping them understand the interactions between complex climate issues and their areas of interest. Web-based map server technology, which has advanced considerably during the past few years, can fulfill these requirements so that the user needs only a web browser and an internet connection to perform analyses.

This proposal addresses the need to make varying-scale historical climate data easily accessible via sophisticated, yet user-friendly, web-based tools that will enable educators and students to incorporate climate information into their teaching and research. To be most effective, these tools must be produced in ways that make them accessible without the need for either specialized software or extensive expertise. We will develop new tools for analyzing climate variability across a wide range of spatial and temporal scales and make the tools freely available as an interactive website and as open source software components.

The project addresses all of NWACC’s key criteria:

INNOVATION:  To develop a web-based tool for analyzing climate variability at multiple scales, we will extend open source internet map server capabilities to include interactive querying and visualization of change over time. GIS-based visualization has proven extremely useful for teasing out patterns in large spatial and temporal datasets. Current web-based map server interfaces typically provide standard tools (zoom, pan, basic attribute query) for navigation and exploration; less frequently, they include tools for generating graphs or animations of time-series data. No support, however, exists for interactively controlling time-series visualizations or querying the underlying data. These new capabilities will be critical in order for educators, students, and others to understand how climate is changing and how interpretive analyses are derived from scientific evidence. The web-based interface will allow users to view animated displays of time-series data and stop animation at any point to query the underlying data at that particular time slice. The results will be presented by visually highlighting those locations in space and time where observed change is considered "atypical," given input from the user and knowledge from experts. The user will then be able to watch the change over time, query the underlying data, and further explore the phenomenon.

Open source GIS and map server software has evolved rapidly in the past few years, providing a rich set of tools for displaying, modifying, and manipulating spatial data. We will leverage existing open source GIS software tools (MapServer, PostgreSQL/PostGIS, GDAL) as the core of the system, developing custom code that interfaces with these tools on the server side and provides an interactive, web-browser-based application on the client side. A significant benefit of these technologies is the absence of any licensing fees (for which commercial equivalents typically cost tens of thousands of dollars), enabling inexpensive implementation of the technology by other educational institutions. Additionally, we will develop our codebase using a common open source programming language (Python) within a structured framework (WSGI, via Pylons) that will facilitate replication at other institutions.

IMPACT:  The web-based toolset will allow advanced exploration of the PRISM climate data. The use of this technology in classrooms and laboratories will enable spatial and temporal climate analysis not previously possible without the use of specialized software tools. Examples of its use include:

  • Undergraduates in GEO 308 (Global Change and Earth Sciences), a core curriculum course at OSU, using historical climate data to examine how patterns of drought have affected the West.
  • Upper undergraduate students in the OSU Geosciences department already proficient in GIS, using historical climate data to identify periods in which recorded precipitation along the precipitation gradient on the west slope of the Cascades varied significantly from "normal." After identifying the periods of interest, the students could download the appropriate datasets for use in their desktop GIS analysis.
  • Graduate or senior-level undergraduate students using the web-based tools interactively to create new web resources about climate variability and its impacts on natural resource management, transportation, or construction techniques in a particular region. The results of these projects could transfer to the K-12 level and other institutions and communities, benefiting students and other interested parties.

FEASIBILITY:  The likelihood of success of this project is extremely high given the resources and experience of the collaborating groups. NACSE (the Northwest Alliance for Computational Science & Engineering) is an interdisciplinary research center widely recognized as the leader in the usability of scientific and engineering applications, and has proven strengths in developing intuitive and highly interactive web-based tools. In particular, Keon and other NACSE staff possess years of experience in the development of web-based mapping interfaces using open source tools. The PRISM group is a leader in producing high-quality spatial climate data sets and analyzing and interpreting climate data to answer real-world questions. The PRISM datasets, which will be used in this project, are the official spatial climate data for the US Department of Agriculture and are widely used by academic and government scientists throughout the world. In addition to the outstanding domain-specific strengths and expertise of the personnel in each group, world-class facilities are available for this project. NACSE and OSU will donate the use of enterprise-level web servers, database servers, fully-backed up disk storage, reliable and fast network connectivity, multiple development platforms (Windows, Linux, Solaris, Mac), and software versioning repositories.

TECHNOLOGY TRANSFER & OUTREACH:  The products of this research will be made available to other institutions and interested parties as a publicly-accessible website for data analysis and exploration and as an open source codebase. Open source software is free and portable, in the sense that it can be obtained by anyone and incorporated into a variety of systems; we will make the codebase available for others to obtain and modify for their own purposes via the open source GPL license. At OSU, we will promote classroom and laboratory use of the web-based toolset across campus, particularly in departments with courses that consider climate variability (e.g., Geosciences, Forest Science, Botany and Plant Pathology, Bioengineering, etc.) or that use PRISM climate datasets for teaching and research. We will also present this work at regional and national conferences, such as GIS in Action and the Association of American Geographers (AAG) annual meeting.

LEVERAGE:  Through additional grants, NACSE will match Keon's programming time 2:1 and PRISM will match Chris Daly’s faculty stipend 3:1, beyond the funding provided by NWACC. NACSE and OSU will donate additional technical support, materials and supplies, hardware use (web & database servers, desktop systems), network access, disk storage, backups, and use of the NACSE usability lab. Additionally, part of this work will be supported by an existing NSF grant award that is funding the development of related tools for examining tsunami disaster scenarios.

COLLABORATION:  The proposed system represents a major step forward in transforming scientific climate data to knowledge that can be directly applied to decision-making tasks. A solution of this type can only be accomplished through the collaboration of groups that already have demonstrated expertise in each of these areas. Our project will leverage the strengths of the NACSE and PRISM groups, who will partner to carry out this project.

OUTCOMES ASSESSMENT:  We will apply usability engineering to this project to eliminate the need for special expertise or training. The completed interface must demonstrate that both students and experts can be exposed to research findings from broader, related areas without requiring too much additional learning curve. Appropriately designed user interfaces provide the features their users require, while ensuring that navigation and use are as "natural" and efficient as possible. Building upon NACSE's experience in usability engineering applied specifically to software, web portal, and shared data systems, we will incorporate established methods as part of prototype development: task-centered design, participatory design, and structured usability testing. The completed interface will be tested by OSU faculty, staff, and students. Usability testing will determine not only how easily the tools can be used by the various groups to find the answers they seek, but also whether the tools function properly across a range of browsers and operating systems.

The development of tools that enable the exploration, analysis, and visualization of climate data will give educators a powerful medium not just for exposing students to the concepts of climate change, but to actually immerse them in the data so that they can generate their own hypotheses and reach their own conclusions. By engaging researchers, educators, and students in the project and applying established usability engineering techniques, we will ensure that the resulting resources and tools will address the needs of real users and serve to promote broader use of science-based climate information in day-to-day decisions.

This project is made possible by support from: