Resource information
Lack of systematic tools and approaches for measuring climate change adaptation limits the measurement of progress toward the adaptation goals of the Paris Agreement. To this end, we piloted a new approach, the Biomass Climate Adaptation Index (Biomass CAI), for measuring agricultural adaptation progress in Ethiopia across multiple scales using satellite remote sensing data. The Biomass CAI can monitor agri-biomass productivity associated with adaptation interventions remotely and facilitate more tailored precision adaptation. The Biomass CAI focuses on decision-support for end-users to ensure that the most effective climate change adaptation investments and interventions can be made in agricultural and food systems. This is the first implementation of such a system in Sub-Saharan Africa. It has not yet been validated and is presented here as preliminary results presenting the concept. This publication has been prepared as an output of CGIAR Research Initiative on Digital Innovation, which researches pathways to accelerate the transformation towards sustainable and inclusive agrifood systems by generating research-based evidence and innovative digital solutions. This publication has not been independently peer-reviewed. Any opinions expressed here belong to the author(s) and are not necessarily representative of or endorsed by CGIAR. In line with principles defined in CGIAR's Open and FAIR Data Assets Policy, this publication is available under a CC BY 4.0 license. © The copyright of this publication is held by IFPRI, in which the Initiative lead resides. We thank all funders who supported this research through their contributions to CGIAR Trust Fund. Methodology:Lack of systematic tools and approaches for measuring climate change adaptation limits the measurement of progress toward the adaptation goals of the Paris Agreement. To this end, we piloted a new approach, the Biomass Climate Adaptation Index (Biomass CAI), for measuring agricultural adaptation progress in Ethiopia across multiple scales using satellite remote sensing data. The Biomass CAI can monitor agri-biomass productivity associated with adaptation interventions remotely and facilitate more tailored precision adaptation. The Biomass CAI focuses on decision-support for end-users to ensure that the most effective climate change adaptation investments and interventions can be made in agricultural and food systems.
A time series of Vegetation Index (VI) specific to locations is modeled for cropland areas. This modeling is based on various external factors like weather elements (such as precipitation and surface temperatures), geographical conditions (including soils, topography, and coordinates), and past NDVI trends. These models produce continuous measures of predicted greenness, allowing for the derivation of predictions for vegetation peak and green-up integrals both spatially and temporally.
The predicted values, calibrated with baseline data, are then compared with the actual values measured by satellites. Significant deviations from the baseline are interpreted as either an enhancement of agricultural status or a deterioration of agricultural activities. These interpretations are viewed through the lens of climate change to identify progress in adaptation and the requirements for intervention.
This is the first implementation of such a system in Sub-Saharan Africa (box -21.463872,17.259292 to 51.880373,3.397636). The data presented here have not yet been thoroughly validated and might change with future recalibration of the models. This dataset is shared as preliminary results. The data shows the average deviation between regression models and the actual satellite measurement within the year 2023. The index theoretical range is between -1 and 1, -1 showing an extreme decrease in agricultural production and 1 showing a strong improvement when compared to the baseline.
For this implementation, the input data are:
MODIS MOD13Q1 for the vegetation status
CHIRPS and GPM data for rainfall precipitation
MODIS MCD12Q1.061 Land Cover Type Yearly Global to identify agricultural lands.
The foundation of the Biomass CAI prediction model is constructed around extensive Convolutional Neural Networks (CNN), which excel in learning intricate and frequently diverse patterns and relationships. CNN proves especially valuable for impact assessments occurring at a specific moment, although it is inherently unsuitable for processing time series data to make immediate predictions using near-real-time input data.
To facilitate the real-time monitoring of climate adaptation progress, we incorporated a Long Short-Term Memory (LSTM) component alongside a CNN. LSTM is renowned for its ability in time series regression, while CNN offers insights into the spatial structure of the agricultural landscape. Through the integration of these components within the workflow, the Biomass CAI model could effectively harness near-real-time data, enabling the timely monitoring of the adaptation process.
For the dataset presented here, the model was calibrated with data from January 2001 to December 2019.
The models and method are presented in more details and a thorough presentation of the concept can be found here:
https://hdl.handle.net/10568/127239
https://hdl.handle.net/10568/135997
https://hdl.handle.net/10568/126091
