Objective/Mission
Overall, our research efforts aim to enable a modeling and analyses framework through which current and future land surface/sub-surface conditions and agricultural productivity under a range of plausible, future climates and changing landscapes are generated, assessed, and disseminated. The purpose of this modeling and analyses framework is to provide projections of trends in key impact variables across agriculture and land systems. Ultimately, we aim to provide actionable information for adaptive measures at regional to local scales for stakeholders, farmers, and communities under threat from changing climate and land conditions.
Approach in Year 3
Our primary modeling tool for this purpose is the Community Terrestrial Systems Model (CTSM, formerly known as the Community Land Model, or CLM). CTSM contains many features relevant to our project research, including detailed process-based biogeophysical and biogeochemical representation of the canopy, surface, and subsurface conditions. CTSM provides explicit modeling of crops (with time-evolving irrigated areas, fertilization, and land use), river modeling, surface/sub-surface hydrology, as well as carbon and nitrogen cycling.
To provide current and future estimates of land conditions, CTSM is driven by climate scenarios generated with the MRCM as well as the MIT Integrated Global Systems Model (IGSM), that includes the MIT Earth-System Model (MESM) as well as the Economic Prediction and Policy Analysis (EPPA) modules. The EPPA economic model has been enhanced under this project to represent the economic sectors and impacts of Bangladesh. It also provides projections of land-use change—but at spatial scales aligned with its larger economic regions. Therefore, to link our econometric-based projections of land-use change to our CTSM spatial configuration, we have enabled a spatial analysis tool, DEMETER, that allows for the spatial interpolation of land-use between models of different spatial resolutions.
Progress in Year 3
Our efforts have been in the construction of a set of “high-resolution” (0.1 degree or ~10 km) CTSM simulations with the observation-customized sowing/planting/transplanting dates of various crop types in Bangladesh. Currently we focus on six crop types, including Boro rice, transplanted Aman rice, mustard, potato, sunflower, and maize. We have completed historical simulations using 3-hourly bias-corrected MRCM atmospheric forcing driven by the ERA5 and CMIP6 EC-Earth3 model. Future simulations are conducted with 3-hourly bias-corrected MRCM forcing driven by the CMIP6 EC-Earth3 model under two SSP scenarios (SSP1-2.6 and SSP5-8.5). Results from the historical simulations are evaluated to assess model fidelity. Future simulations are used for analyses of trends/changes in key diagnostics (e.g., water logging, drought, crop yields, etc.).
As part of Jameel Observatory CREWSnet’s “advanced crop calendar,” we have focused our analyses on the diagnoses of crop hazards to waterlogging and (agricultural) drought conditions. The assessment of “waterlogging” (i.e., saturated conditions that last at least three days) has extracted the frequency and duration of surfacing ponding and root-zone saturation over cultivated areas – and consider both irrigated and rainfed crops (see Figure 13). Our drought assessment focuses on agricultural drought, namely various impacts (soil water deficit, reduction of evapotranspiration, reduced crop yield, etc.) of meteorological drought on agriculture, but will extend to other types of droughts (e.g., hydrological drought, socio-economical drought).
Foundational/internal modeling
Initial results from the waterlogging assessment indicate widespread changes in frequency and duration of waterlogging (across both crop types and location) as well as the potential for more “extreme” episodes from year to year (e.g.,Figure 13). We have employed one of the widely used drought indices, Standardized precipitation and evapotranspiration index (SPEI). Different drought characteristics (duration, severity, intensity, and frequency) are being evaluated at a range of time scales (1-, 2, 3-, 6-, 12-, 24-, 36-, and 48-month) to convey the full scope of this hazard.
The waterlogging and drought assessment will, in part, form the basis of an interactive, use-inspired visualization platform to display various crop hazards that ultimately will support the development of Jameel Observatory CREWSnet’s Advanced Crop Calendar. We will work with our BRAC colleagues to undertake key engagement activities with stakeholder and communities to determine the most effective visualization.
Data Hub and Decision-Support Tool
Our simulations and analyses of waterlogging and drought risks for the variety of crop types will be made publicly available and accessible via the Jameel Observatory CREWSNet data hub portal. Working with our BRAC colleagues, we will engage with regional to local stakeholders and farmers to present these changing risks through the Jameel Observatory CREWSnet decision-support tool and visualization platform.