Periodic Reporting for period 2 - CONFER (Co-production of Climate Services for East Africa)
Berichtszeitraum: 2022-03-01 bis 2023-08-31
CONFER represents a collaborative, multinational effort dedicated to strengthening resilience against climate-related impacts and reducing disaster risk in East Africa. This region has experienced significant vulnerability to climate-related shocks, as evidenced by recent extended periods of droughts and severe flooding, which have left millions in need of assistance. Our primary objective is to enhance the quality of life in the region by co-creating specialized climate services tailored to the water, energy, and food security sectors. We achieve this by working closely with a diverse array of stakeholders and end-users to bolster their ability to plan for and adapt to seasonal climate variations.
The scope of CONFER is vast, encompassing eleven countries and approximately 365 million people. Our scientific endeavors are ambitious and revolve around three interconnected tracks:
End-User Engagement: We actively engage with end-users through the Greater Horn of Africa Climate Outlook Forum, which convene three times a year and draw in numerous stakeholders. These platforms serve as spaces for the collaborative development of novel, sector-specific climate services. By fostering two-way communication between our scientific experts and a broad spectrum of stakeholders and end-users, we aim to generate enthusiasm and raise awareness, ensuring that the value of our new scientific innovations and products is fully realized by those who stand to benefit the most.
Research Advancements: Our research efforts focus on enhancing the accuracy and localized detail of numerical prediction model outputs for East Africa, with particular emphasis on seasonal prediction. These advancements have specific applications in the fields of hydrology and crop modelling.
Statistical and Machine Learning Tools: We are at the forefront of developing statistical and machine learning tools to attain a heightened level of seasonal forecast accuracy.
Our team of scientific experts is deeply involved in an extensive training and capacity development program aimed at bolstering the uptake of climate information within our focus sectors. Our ultimate goal is to equip the region with the knowledge and tools needed to adapt to climate challenges.
The scope of CONFER is vast, encompassing eleven countries and approximately 365 million people. Our scientific endeavors are ambitious and revolve around three interconnected tracks:
End-User Engagement: We actively engage with end-users through the Greater Horn of Africa Climate Outlook Forum, which convene three times a year and draw in numerous stakeholders. These platforms serve as spaces for the collaborative development of novel, sector-specific climate services. By fostering two-way communication between our scientific experts and a broad spectrum of stakeholders and end-users, we aim to generate enthusiasm and raise awareness, ensuring that the value of our new scientific innovations and products is fully realized by those who stand to benefit the most.
Research Advancements: Our research efforts focus on enhancing the accuracy and localized detail of numerical prediction model outputs for East Africa, with particular emphasis on seasonal prediction. These advancements have specific applications in the fields of hydrology and crop modelling.
Statistical and Machine Learning Tools: We are at the forefront of developing statistical and machine learning tools to attain a heightened level of seasonal forecast accuracy.
Our team of scientific experts is deeply involved in an extensive training and capacity development program aimed at bolstering the uptake of climate information within our focus sectors. Our ultimate goal is to equip the region with the knowledge and tools needed to adapt to climate challenges.
Our work is now primarily structured within interdisciplinary teams referred to as "workstreams." The first three workstreams focus on enhancing existing products and services, with an emphasis on probabilistic forecasts for seasonal and sub-seasonal timescales. Additionally, efforts are directed toward improving early actions and anticipatory responses by the Kenya Red Cross.
Furthermore, three workstreams are dedicated to the development of new climate services. The first of these focuses on establishing a forecasting system for streamflow and flooding in flood risk hotspots. In this regard, we have been actively working on operationalizing a regional setup of WRF-Hydro for the entire region to provide weekly forecasts. Additionally, we have been developing a prototype seasonal hydrological forecast for a case study in the Tana basin, particularly for the OND seasons in 2018 and 2019.
The second new service revolves around the improvement of forecasting the onset and cessation of rainy seasons, which are crucial aspects identified by end-users. Our work has centred on developing a new and consolidated definition of the onset.
The third newly developed service focuses on enhancing crop modelling and forecasting for plant health and yield. Thus far, this work has led to the creation of a joint croplands and rangelands product that performs exceptionally well in terms of spatial accuracy and captures a significant portion of inter-annual variability.
The final workstream aims to assess the information flow within the value chain for climate services in East Africa.
Moreover, we have actively engaged in co-production activities, including supporting the preparation and execution of co-production sessions at four Greater Horn of Africa Climate Outlook Forum (GHACOF) events. Training has been a pivotal undertaking with substantial outcomes. We began by organizing interviews, discussions, and surveys involving various stakeholders and beneficiaries to determine the support needed to deliver effective climate services. The results have informed a training plan covering key areas and have been instrumental in shaping training workshops and collaborations with projects beyond CONFER, such as WISER.
Our team has also facilitated the delivery of three-week, digital annual seasonal climate prediction workshops, in addition to pre-GHACOF capacity-building workshops.
Furthermore, three workstreams are dedicated to the development of new climate services. The first of these focuses on establishing a forecasting system for streamflow and flooding in flood risk hotspots. In this regard, we have been actively working on operationalizing a regional setup of WRF-Hydro for the entire region to provide weekly forecasts. Additionally, we have been developing a prototype seasonal hydrological forecast for a case study in the Tana basin, particularly for the OND seasons in 2018 and 2019.
The second new service revolves around the improvement of forecasting the onset and cessation of rainy seasons, which are crucial aspects identified by end-users. Our work has centred on developing a new and consolidated definition of the onset.
The third newly developed service focuses on enhancing crop modelling and forecasting for plant health and yield. Thus far, this work has led to the creation of a joint croplands and rangelands product that performs exceptionally well in terms of spatial accuracy and captures a significant portion of inter-annual variability.
The final workstream aims to assess the information flow within the value chain for climate services in East Africa.
Moreover, we have actively engaged in co-production activities, including supporting the preparation and execution of co-production sessions at four Greater Horn of Africa Climate Outlook Forum (GHACOF) events. Training has been a pivotal undertaking with substantial outcomes. We began by organizing interviews, discussions, and surveys involving various stakeholders and beneficiaries to determine the support needed to deliver effective climate services. The results have informed a training plan covering key areas and have been instrumental in shaping training workshops and collaborations with projects beyond CONFER, such as WISER.
Our team has also facilitated the delivery of three-week, digital annual seasonal climate prediction workshops, in addition to pre-GHACOF capacity-building workshops.
We have undertaken several research endeavours to deepen our understanding of climate prediction and its impact. In a 2021 paper we delved into the relationship between initial-state sea surface temperatures and subseasonal rainfall forecast errors using a cutting-edge forecast model. Our findings highlighted that the initial mode of the Indian Ocean Dipole (IOD) partially controls rainfall errors in weeks 3-4. The model exhibited overly wet forecasts in 2015 when the IOD was positive and excessively dry forecasts in 2010 during a negative IOD phase.
In a subsequent 2022 paper, we sought to unravel the strengths and weaknesses of statistical prediction models. Our research revealed that the August states of El Niño-Southern Oscillation (ENSO) and the Indian Ocean Dipole (IOD) could predict nearly half of the East African short rains variability between 1950 and 2020. We also explored the potential benefits of upfront adjustments based on initial conditions, noting cases where they could have improved forecasts and instances where they would have exacerbated predictions.
A third paper delved into the consequences of afforestation in a heavily farmed tropical African region. The study revealed that afforestation led to more precipitation and runoff. This discovery has significant implications for regional water resources and climate policy decisions.
Another CONFER paper, focused on enhancing seasonal forecasts for the Greater Horn of Africa (GHA). The research explored the benefits of informed subselection of ensemble members from global models in comparison to random subselection. Emphasizing seasonal rainfall predictions over the GHA, the results consistently showed that informed subselection outperformed random selection, especially with small ensemble sizes and during seasons with active teleconnections. These techniques are not only feasible for operational use but also contribute to enhanced forecast accuracy.
In a 2023 paper we explored soil moisture characteristics in Ethiopia and their connection to local and remote factors. Using data analysis techniques, we identified regional patterns, with northeastern Ethiopia experiencing consistent dry conditions, and the western region exhibiting a persistent wetter trend.
In the remaining phase of the project, we anticipate advancing the state of climate prediction for East Africa across various domains, including dynamical downscaling, hydrological modelling, statistical post-processing, rainy season onset prediction, and crop modelling. These achievements, in conjunction with our comprehensive training program and communication and dissemination efforts, are poised to yield substantial societal benefits by ensuring higher-quality climate services for stakeholders and decision-makers in East Africa.
In a subsequent 2022 paper, we sought to unravel the strengths and weaknesses of statistical prediction models. Our research revealed that the August states of El Niño-Southern Oscillation (ENSO) and the Indian Ocean Dipole (IOD) could predict nearly half of the East African short rains variability between 1950 and 2020. We also explored the potential benefits of upfront adjustments based on initial conditions, noting cases where they could have improved forecasts and instances where they would have exacerbated predictions.
A third paper delved into the consequences of afforestation in a heavily farmed tropical African region. The study revealed that afforestation led to more precipitation and runoff. This discovery has significant implications for regional water resources and climate policy decisions.
Another CONFER paper, focused on enhancing seasonal forecasts for the Greater Horn of Africa (GHA). The research explored the benefits of informed subselection of ensemble members from global models in comparison to random subselection. Emphasizing seasonal rainfall predictions over the GHA, the results consistently showed that informed subselection outperformed random selection, especially with small ensemble sizes and during seasons with active teleconnections. These techniques are not only feasible for operational use but also contribute to enhanced forecast accuracy.
In a 2023 paper we explored soil moisture characteristics in Ethiopia and their connection to local and remote factors. Using data analysis techniques, we identified regional patterns, with northeastern Ethiopia experiencing consistent dry conditions, and the western region exhibiting a persistent wetter trend.
In the remaining phase of the project, we anticipate advancing the state of climate prediction for East Africa across various domains, including dynamical downscaling, hydrological modelling, statistical post-processing, rainy season onset prediction, and crop modelling. These achievements, in conjunction with our comprehensive training program and communication and dissemination efforts, are poised to yield substantial societal benefits by ensuring higher-quality climate services for stakeholders and decision-makers in East Africa.