Clark Center for Geospatial Analytics (Clark CGA) was established in 2023 under the leadership of Director Prof. Hamed Alemohammad. Clark CGA’s mission is focused on conducting transformative research using geospatial analytics to address global environmental challenges, with an emphasis on climate change adaptation and mitigation, ecologic conservation, and agriculture. Committed to an entrepreneurial role in technology, application, and accessibility, Clark CGA aims to make lasting contributions to a sustainable future through collaborative, interdisciplinary efforts, marking a continuation of Clark Labs’ legacy within Clark CGA’s ambitious mission. Clark CGA integrated Clark Labs’ activities and team, that are now part of the greater Clark CGA institution and mission.
Clark Labs was founded in 1987 by Prof. Emeritus J. Ronald Eastman, and pioneered IDRISI, the first GIS software for microcomputers which quickly gained traction in academia and the international development community. Throughout the 1990s, Clark Labs expanded into developing decision support tools for GIS, focusing on sustainable development and conservation. Key developments included the introduction of the Land Change Modeler and Earth Trends Modeler. Collaborations with organizations like USAID, WWF, and the Gordon and Betty Moore Foundation led to innovative tools for land cover mapping and supporting REDD projects. Clark Labs also introduced multi-criteria decision-making tools to GIS, transforming the field and leaving a legacy of groundbreaking research and strategic partnerships.
A detailed history of Clark Labs is described below:
The development of IDRISI and TERRSET
Founded by Emeritus Professor of Geography J. Ronald Eastman in 1987, Clark Labs was initially established to maintain the development and technical support of the IDRISI GIS and Image Processing System. IDRISI was the first GIS written expressly for a microcomputer platform, with the goal of providing a low-cost modular system that could:
- Function without costly additional hardware such as expensive non-standard graphical display systems (as was the norm at the time),
- Process large data sets with minimal RAM (random access memory),
- Allow for user-extensions through the use of an open architecture and intentionally simple file formats, and thereby
- Foster the analytical development of GIS.
The earliest versions supported CP/M, MS-DOS and Vax-VMS on systems such as the DEC Rainbow, IBM PC and DEC Micro-VAX. However, as the microcomputer industry quickly developed, this was narrowed to focus on MS-DOS.
The late 1980s was a time of rapid development in GIS, paralleling the success of microcomputers. In 1988, the National Center of Geographic Information and Analysis (NCGIA) was established with funding from the National Science Foundation. Among its first objectives was to create a national curriculum in GIS. To support this, the NCGIA selected IDRISI as the system for the tutorials because of its accessibility and rapid adoption by universities. Also in the late 1980’s, the United Nations Environment Program (UNEP) selected IDRISI to support the implementation phase of its program of technology transfer in GIS funded by the Swiss Agency for Development and Cooperation and supported by the United Nations Institute for Training and Research (UNITAR). A key element for UNEP and UNITAR was support for the processing of remotely sensed images on the new IBM PS/2 computers with XGA displays gifted by IBM to support the program. Accordingly, Clark Labs quickly incorporated a suite of image processing tools, with financial support from UNEP. From 1989-1994, Clark Labs also provided training services for UNEP/UNITAR is locations across Africa, Eastern Europe, South America and Asia.
In the 1990’s Clark Labs’ strong connection to international development applications of GIS and Image Processing continued, especially with USAID. The largest of these was a partnership with the University of Arizona to implement an environmental monitoring program in Malawi, funded by USAID.
Despite a general sense in the 1990’s that GIS was an important tool for informing decision making, experience working in developing nations and witnessing the challenges of natural resource development quickly led the Clark Labs team to realize that decision support tools in GIS were severely lacking. This led to the first truly distinctive research contribution, to develop a suite of multi-criteria/multi-objective decision support tools.
Another consequence of its work in international development was a realization of the challenges of sustainable development. As a result, Clark Labs strongly oriented the focus of IDRISI to support, in particular, the conservation community. This led to a long history of special partnerships with the Gordon and Betty Moore Foundation, Conservation International and the Wildlife Conservation Society. Following from this, Clark Labs also developed a strong focus on empirical modeling tools including an innovative suite of neural network tools, which led, in turn, to the development of substantial vertical applications such as the Land Change Modeler, the Earth Trends Modeler, the Habitat and Biodiversity Modeler, the Climate Change Adaptation Modeler and the Ecosystem Services Modeler. In 2014, this full suite of tools was renamed TerrSet, with IDRISI still providing the GIS and Image Processing backbone.
Clark Labs research history
In addition to geospatial software development, Clark Labs has also had a long history of applications research and close partnerships with organizations such as the Africa Bureau of the United States Agency for International Development (USAID), the United Nations Environment Program Global Resource Information Database (UNEP-GRID), the United Nations Institute for Training and Research (UNITAR), Conservation International, the Gordon and Betty Moore Foundation, the Wildlife Conservation Society, the US Department of Agriculture’s Animal Plant Health Inspection Service (USDA-APHIS) and NASA. Key areas of research include the following:
Land change modeling
In 2003, Conservation International contracted with Clark Labs to develop a predictive land change modeling capability. This led to the development of the Land Change Modeler (LCM) and its associated Habitat and Biodiversity Modeler (HBM) for the conservation assessment of land change impacts. As of early 2024, over 3300 scholarly publications featuring predictive land change modeling used LCM. Particularly attractive was the availability of a wide range of statistical and machine learning tools for empirical modeling, including a Multi-Layer Perceptron neural network, Decision Forest (an implementation of Random Forest), Logistic Regression and a Support Vector Machine.
Building on the knowledge gained with LCM, Clark Labs partnered with Esri to develop a cloud-based implementation of LCM specifically for the United States at 30 m resolution. The American Land Change Modeler (ALCM) was released in 2018, and provides a prediction of land cover change in the US to 2050. Esri and Clark Labs have continued their partnership in this arena. In 2021, Clark Labs developed the Esri Land Cover 2050 global land cover change prediction for the mid-century at 300 m resolution.
Time series analysis
The partnership between Clark Labs and the Gordon and Betty Moore Foundation began in 2006 with a grant to develop the Earth Trends Modeler (ETM). The original intention of ETM was to monitor spatial-temporal trends in biodiversity indicators using earth observation data. This quickly grew to add a comprehensive suite of tools for image time series pattern analysis, including T-mode and S-mode PCA, Multichannel Singular Spectrum Analysis, Canonical Correlation Analysis and Empirical Orthogonal Teleconnection (EOT). ETM and LCM are the most distinctive research projects undertaken by Clark Labs.
REDD support tools
REDD (Reducing Emissions from Deforestation and Forest Degradation) is an important greenhouse gas emission reduction strategy. In support of this, Clark Labs added a REDD analysis facility to the Land Change Modeler, based on the World Bank’s BioCarbon Fund (BioCF) Methodology for Estimating Reductions of GHG Emissions from Mosaic Deforestation. Building on this, Conservational International and Clark Labs worked together to develop GEOSIRIS – a national level REDD planning tool, implemented in TerrSet. Clark Labs also partnered with Google to implement a REDD-focused tool using the earliest implementation of Google Earth Engine. More recently, Clark Labs has been working with Verra (the leading carbon registry for the voluntary-sector REDD projects) to develop its VT0007 Unplanned Deforestation Allocation (UDef-A) tool.
Land/forest cover mapping
In the development of image processing tools in IDRISI/TerrSet, Clark Labs developed many unique tools, particularly with respect to image classification (semantic segmentation). TerrSet includes the most extensive set of soft classifiers in the industry, many of them unique in the geospatial community. One of these, the Mahalanobis Typicality classifier, provided the basis for a multi-year research project with the USDA Animal Plant Health Inspection Service (APHIS) to try to detect trees infested with Emerald Ash Borer. It also provided the basis for detection of unknown forest communities in the Calakmul Biosphere Reserve as part of a multi-year forest and land cover mapping in the Yucatan Peninsula, funded by NASA.
From 2013-2024, Clark Labs has worked with the Gordon and Betty Moore Foundation to map mangroves and pond aquaculture in 17 countries across the tropics at 15 m resolution. Maps have been produced for 1999, 2014, 2018, 2020 and 2022 in support of the foundation’s work to foster sustainable shrimp production. Key partners in this work include the World Wildlife Fund and the Aquaculture Stewardship Council.
Multi-criteria/multi-objective decision making and empirical modeling
In 1993, Clark Labs introduced the first instance of Multi-Criteria and Multi-Objective decision making tools in GIS. Innovations include the first implementation of the MOLA heuristic for multi-objective land allocation which also underlies the mechanics of the Land Change Modeler; the first implementation of the Ordered-Weighted Average for multi-criteria evaluation that allows one to balance the relative amount of tradeoff between criteria with decision risk in balancing discordant information; and the first GIS software implementation of Saaty’s Analytical Hierarchy Process (AHP). These methods became important tools in the work of the Malawi Environmental Monitoring Program (MEMP) conducted in partnership with the University of Arizona and funded by USAID.
Building on these tools, Clark Labs has placed strong emphasis on the development of empirical modeling tools, including a distinctive set based on neural networks (MLP, SOM, Fuzzy Artmap). An early application of these was a project to predictively map the interstate spread of Gypsy Moths via containers and shipments for the USDA Animal Plant Health Inspection Service.