Research accelerator project on future earth observation systems for drought management

The project outcomes are:

1. Assess a new open-source dataset to be integrated in DEWS-NDMA and IBLI-ZepRE operational systems and its comparison against currently used datasets that supports the continuity of operational programs; (short term outcome)
2. Development of a processing pipeline for innovative drought indices derived from high-resolution operational datasets such as Sentinel-2 that supports more informed decisions about the development of the next-generation drought indices; (medium-term outcome)
3. Establishment of a community of practice on drought index design in Kenya that facilitates the integration and harmonization of different anticipatory drought risk management and financing initiative.

The project targets institutions and private sector actors involved in the development, implementation and evaluation of tools and initiatives, which are aimed at improving drought risk management through anticipatory action and rapid financial response.

Through the activities, N’GOMbE aims to impact practices and capacities on anticipatory drought risk management:

1. The co-design and testing of new operational datasets improves the capacities of relevant organizations of technical evaluation of alternative EO datasets and drought indices. This is an important step for standardization of quality standards for risk management and financing instruments.
2. The research on innovative high-resolution indices promotes a closer collaboration between institutional stakeholders, private sector actors and knowledge institutions (research centres, local and international universities) on the improvement and innovation of operational systems through evidence-based research.
3. The establishment of the community of practices creates the enabling conditions for stronger cooperation among private and public sector actors involved in drought risk management initiatives and facilitate cost-effective investments on drought index design.

Overall, a strong and sustained collaborative agenda on the design of drought indices is expected to improve the quality and impact of initiatives that trigger anticipatory response mechanisms, thus making them more effective and beneficial for the target pastoral communities.

The project is organized in three Work Packages:

WP1. We will test a new dataset (“MODIS-VIIRS”) developed by scientists at the Joint Research Centre (JRC). It provides global harmonized fAPAR 10-days composite time series from MODIS and VIIRS instruments from 2000 onward at 500 m spatial resolution. First, we will collect and organize all the required satellite and ancillary (index-units, rangeland masks, ground datasets etc.) datasets from JRC and local partners. Then, we will develop a processing chain for the MODIS-VIIRS dataset following the existing operational methodology used for the drought indices of DEWS and IBLI. Finally, the drought indices calculated using the new dataset will be evaluated against current operational indices and the results reviewed with NDMA, ZEP-RE and partners. To facilitate the direct use of the MODIS-VIIRS dataset, we will develop an open tool (R package) to calculate the drought indices following the DEWS and IBLI methodology.

WP2. We will evaluate the potential of fine resolution open-access satellite datasets such as S2 (10m) and Planet NICFI (5m) to calculate drought indices for early-warning and insurance. The main challenge with S2 images is their limited temporal resolution (3-5 days) that, especially in the tropics, could be a constraint for monitoring intra-seasonal vegetation changes. Planet NICFI is already distributed as a monthly composite product, while for S2 a processing algorithm will be developed to produce monthly composite images. After data acquisition and compositing, the data processing will follow the same steps as for the operational coarse resolution indices. The fine-resolution indices will be compared with the ones generated in WP1 for selected geographic locations with good data availability (including ground data), such as Samburu or Marsabit Counties. The results will be reviewed with NDMA, ZEP-RE and other stakeholders to evaluate opportunities and limitations of using fine resolution data for the next generation of anticipatory DRMF systems.

WP3. This cross-cutting WP will make sure that a co-learning/co-creation infrastructure will be in place during and after the project duration. This task will be achieved by organizing 1 inception meeting, 2 workshops and 2 webinars during the project and by designing a roadmap toward the establishment of a community of practice aimed at stimulating a continuous cross-learning and open debate on how to design more effective and better harmonized EO-based anticipatory drought risk management tools in pastoral Africa. As part of the roadmap, priority EO research areas will be identified to guide future longer-term projects and research activities.

The following key outputs are expected:

1. Quality assessment of a new open long-term dataset for drought monitoring and open-source tools to facilitate its use for DEWS, IBLI and other programs.
2. Design and evaluation of innovative drought indices at fine spatial resolution. This output will inform decisions on future index design.
3. Establishment of a community of practice that can facilitate a more effective and impactful use of EO data for drought early warning and risk financing in Kenya and beyond.

N’GOMbE pillars to reach short and longer-term outcomes are:

1. Address concrete needs through evidence-based research: NGOMbE delivers solutions that could be integrated into operational early warning and index-insurance initiatives and that contribute to their effectiveness. Open-source datasets and tools will be also beneficial in terms of reduced implementation costs and increased transparency.
2. Co-design the project with local actors: NGOMbE promotes capacity development throughout the project implementation. Specifically, we will support local actors in building their capacity in evaluating alternative data products and in assessing trade-offs between EO data sources/processing methods for operational programs.
3. Promote long term dialogue, cross-learning, and coordination: through the creation of a community of practice, NGOMbE will create the enabling conditions for a better technical coordination and harmonization of existing DRMF instruments in Kenya, thus improving their efficacy for crisis prevention.
4. Facilitate synergies with ongoing projects in pastoral drylands: NGOMbE partners are involved in several large projects targeting pastoral communities in East Africa, such as Resilient Approaches in Natural ranGeland Ecosystems (RANGE), De-risking, Inclusion and Value Enhancement of Pastoral Economies in the Horn of Africa, the CGIAR Livestock and Climate Initiative. This connects NGOMbE with broader action research agenda with complementary interventions that would reinforce the value of our results.

The partnership is built to support these pillars. While all partners are involved from the beginning in the co-design of the project, with local partners expressing clear needs for operational systems, each will play a specific role in the impact pathways:

• The teams of the University of Milan and University of Twente have a long history of applied research in East Africa on EO applications for drought early warning and risk financing instruments. They will guarantee that the generated outputs will be tailored for the specific context while meeting high scientific quality standards.
• ILRI (Kenya) is ideally positioned to act as catalyser for the innovative solutions proposed in the project and can play an important role in the broader communication and regional scaling of the project findings. ILRI could act also as a convener of the community of practice, facilitating a long-term engagement among national and regional institutions.
• NDMA (Kenya) and ZepRE (Kenya) are two critical national and regional players in the implementation of anticipatory risk management systems and manage two large operational initiatives. They will act as early adopters of the proposed solutions and contribute to their assessment and potential operationalization in their systems.

The project is led by the Università degli Studi di Milano, with Zep-RE PTA Reinsurance Company, Kenya; National Drought Management Authority, Kenya; Seidor Consulting (JRC), Italy; International Livestock Research Institute, Kenya; and University of Twente, The Netherlands.