CGIAR Geospatial Strategy

CGIAR Geospatial Strategy proposes a pathway for CGIAR to effectively leverage geospatial technology to support its reaffirmed mission of delivering "science and innovation that advance the transformation of food, land, and water systems in a climate crisis." This document includes considerations of CGIAR's geospatial analysis capacity, data management, research approaches, and technology that are collectively needed to achieve CGIAR’s Collective Global 2030 Targets.

DRAFT v1.0 | DECEMBER 16, 2021

Prepared by Jawoo Koo (IFPRI/Big Data) with inputs from Elliott Dossou-Yovo (AfricaRice), Hannes Gaisberger (Bioversity Int’l), Steven Prager (CIAT), David Gaveau (CIFOR), Kai Sonder (CIMMYT), Henry Juarez (CIP), Chandrashekhar Biradar (ICARDA), Muhammad Ahmad (ICRAF), Murali Krishna Gumma (ICRISAT), Zhe Guo (IFPRI), Tunrayo Alabi (IITA), Francesco Fava (ILRI), Renaud Mathieu (IRRI), Kiran Chandrasekharan (IWMI), Shwu Jiau Teoh (WorldFish), and Andy Nelson (U. Twente/ITC)

Why does CGIAR need a geospatial strategy?

The rapid development of high resolution, high-frequency satellite remote sensing unlocks the unprecedented capability to monitor food, land, and water systems precisely and timely at scale. CGIAR’s 2030 Research and Innovation Strategy presents an ambitious roadmap of how CGIAR will develop and deploy its capacities, assets, and skills to address priority global and regional challenges with partners over the decade. The Strategy presents “Make the digital revolution central to our way of working” as one of seven implementation approaches and recognizes the role of big data and Earth observation to help achieve impacts at scale. There are, however, unaddressed technical and organizational challenges that need to be overcome to realize the full potential of geospatial data analytics. To this end, we, as the geospatial science CoP, call for collective action to close the gap.

Strategic Goals

1. Basemaps
Provide the best-in-class geospatial data and analysis tools to CGIAR and partners.

2. Groundtruthing data
Improve data sharing and technical collaboration between CGIAR and partners.

3. Remote sensing capacity
Enhance satellite remote sensing analysis and in-season analytics capacities.

4. Shared services
Provide enabling Shared Services to accelerate the adoption of new technologies.

5. Data science capacity
Mainstream data science approaches to ensure reproducibility, support policy decisions, and deliver impacts.


  1. Investment in foundational geospatial datasets has been inadequate. Outdated baseline datasets are used in targeting, training, and monitoring, and evaluation.
  2. Groundtruthing and training data are not strategically collected and shared. Publicly available datasets are inconsistent and underrepresenting small-scale producers’ farming systems.
  3. Satellite remote sensing analysis capacity is limited. Most of the geospatial capability focuses on the re-analysis of secondary datasets. Demands for applying satellite remote sensing data for monitoring food, land, and water systems are unmet.
  4. Geospatial science relies on many external data sources and tools. Duplicated efforts exist to engage with technical partners and service providers. Fair use of shared resources requires careful management and discipline.
  5. Data science skills are highly uneven. The low availability of computer programming and statistical learning talents becomes a bottleneck to provide timely analytics, participate in external initiatives, access third-party data sources.

Proposed Plans

  1. Basemaps
    • Assess the critical gaps in the commonly used foundational global baseline datasets
    • Identify internal or external capabilities to develop, update, maintain the datasets
    • Embed the provision of baseline geospatial data in Impact Area Platforms with dedicated resources and long-term commitment.
  2. Groundtruthing data
    • Adopt the standard of groundtruthing data collection, management, and sharing, following the FAIR principle.
    • Identify benchmark locations and key data types to meet Initiatives‘ needs and support partners’ policy decisions for impacts.
    • Embed the provision of groundtruthing data as a key digital service activity.
  3. Remote sensing capacity
    • Assess the current capacity and skill levels of CGIAR scientists to analyze satellite remote sensing to generate real-time insights to meet Initiatives’ needs and expectations.
    • Establish technical relationships with external partner institutions and initiatives, leveraging CGIAR’s global reach for providing groundtruthing data, domain expertise, local knowledge, and direct development impact opportunity.
    • Provide cloud-based storage and compute environment to empower CGIAR scientists to analyze satellite imagery data without being constrained by local IT infrastructure.
  4. Shared services
    • Assess the demand for useful geospatial data and enabling subscription-based services (e.g., high-frequency satellite remote sensing, online analysis, and visualization tools, lease of UAV equipment and piloting services, groundtruthing data collection sensors).
    • Establish an institutional Shared Services unit in Digital Services to facilitate procurement, service provision, manage usage, and coordinate training and helpdesk activities.
    • Actively engage with the FOSS4G (Free and Open Source Software for Geospatial) community to improve the accessibility and applicability of open source tools in collaborative research with partners.
  5. Data science capacity
    • Cultivate data and code sharing culture to increase synergies, reduce duplicative data engineering efforts, and improve geospatial analysis reproducibility across CGIAR.
    • Provide on-demand online training that guides CGIAR scientists to use computer programming-based geospatial data analysis, develop machine learning applications, and participate in guided projects to familiarize new geospatial data and technologies.
    • Adopt a decentralized geospatial data science organizational structure that geospatial analysts are strategically located to best support Initiatives while utilizing shared analytical infrastructure and adopting standardized practices for analytics and documentation across CGIAR.

Prioritized Activities in 2021

  1. Assessment of CGIAR’s geospatial capacity status and future needs
    • Talents: Assess internal expertise and capacity on geospatial analysis-related skills (e.g., GIS, satellite remote sensing, UAVs, machine learning, computer programming, visualization) and domain knowledge (e.g., crop mapping, natural resource management, digital soil mapping, land use/change mapping, and socioeconomic indicators).
    • Baseline data: Assess key foundational baseline datasets that are widely used (e.g., population, crop type, crop productivity, land use, rural poverty, market access, value-chain) and identify opportunities to improve their relevance to CGIAR’s research.
    • Groundtruthing data: Assess ongoing groundtruthing data collection activities and identify opportunities to improve their value and utility in shared data science applications.
    • Initiatives: Assess the portfolio of One CGIAR Initiatives, in the context of CGIAR 2030 Research and Innovation Strategy, to identify their geospatial data and analysis capacity needs.
    • Technology: Assess the currently available geospatial data and analytical infrastructure across CGIAR and identify opportunities to improve collectively under One CGIAR.
  2. Strategic engagement with One CGIAR’s Design Working Groups and Science Groups
    • Digital Services, in relation to:
      1. Management of Shared Services for geospatial data, software, and subscription services
      2. Provision of shared analytical infrastructure to support One CGIAR Initiatives
      3. Establishing a centralized data and code repository
      4. Standardizing practices for geospatial data management, analytics, documentation, and responsible management of potentially sensitive research data with geospatial information.
    • People and Culture, in relation to:
      1. Decentralized management of geospatial data scientists
      2. Strategy to retain highly-skilled talents in geospatial science domain
      3. Reskilling CGIAR scientists with advanced geospatial technologies
    • Impact Area Platforms, in relation to:
      1. Investment on key geospatial baseline datasets
      2. Scoping supports to the thematic Communities of Practice, including geospatial data and analysis (i.e., CGIAR-CSI).
    • Science Groups, in relation to:
      1. Strategy to leverage geospatial data and Earth observation technology in the portfolio of forthcoming One CGIAR Initiatives
      2. Coordinate to establish (or reaffirm) strategic partnerships with academic institutions (for capacity building), technical partners (for data sharing and innovation), and regional/national stakeholders (for decision support and policy impacts).