Short overview

The Climate, Land, Energy, Water systems (CLEWs) methodology is a model-based methodology to assess costs and benefits of policy and investment decisions made in one sector (e.g. land use) on the other sectors (e.g. water supply). CLEWs models can be developed with different approaches and different modelling tools. One of them, largely employed by UNDP and UNDESA in capacity development activities worldwide, uses the Open Source Modelling Tool (OSeMOSYS), a technologically-detailed techno-economic optimisation tool. By using this tool, we create a techno-economic representation of the climate, land, energy and water systems, where the parts of these systems are represented as a number of processes with certain transfer functions and exchanging between them different commodities. For example, one particular agricultural land use is represented as a ‘box’ that takes a certain quantity of water, energy and land area as inputs and delivers part of the water (through ground water recharge or surface run off) and a crop with a certain yields. In this way, the total land use of the region can be represented through many such ‘boxes’ in parallel, with different inputs and different yields, and a competition between these uses can be created, on the basis of their costs and the resource availability. The optimisation seeks to minimise the Net Present Value of all costs incurred across the water, energy and land sectors in the whole time domain analysed (typically, of several decades), while meeting an increasing demand for commodities (e.g. food products) and resource availability constraints.

The time domain of an application may cover e.g. up to 2070 or more. A good time resolution that allows rather detailed representation of renewables without making the computation incredibly heavy could be around 100 time slices. The energy sector is the one requiring the highest time resolution, so such time resolution would be good also for the other sectors. The energy system typically could include any energy conversion and energy storage process like other TIMES applications. It can be regionalized and some power infrastructure can be represented individually, if important. Any type of emission can be represented, as long as emission factors are available and as long as emissions are linearly dependent on generation. The land system model represents many possible types of land uses, including built-up land, forest, grassland, barren land, water bodies, other (e.g. natural reserves), pastures and crops. Crops are further divided in any type of crop that is relevant in the region being analysed. Each crop is again divided in 5 categories of land uses, depending on the type of inputs: irrigated land with intermediate or high input (in terms of fertilizers, chemicals, mechanization and energy) and rainfed land with low, intermediate and high input. Each of these types will have different yields and they compete between themselves in the model, based on cost- and resource-optimality criteria. Finally, land uses are also divided in clusters, which aggregate cells of the region with similar agro-climatic conditions. All this categorization and all input data are usually derived from GAEZ, but data can (and should) be extracted from local/national sources, if they are available. Emissions from land use changes can also be represented, e.g. following the standards defined by IPCC for emissions accounting. The water system represents precipitation, all water inputs and outputs of land uses, and all final water demands / uses. Precipitation, evaporation from water bodies and evapotranspiration from vegetation are represented. They vary according to climatic conditions and climate change. Groundwater and surface water uses are represented separately (but need improvement). The water system is part of the optimisation. However, it can also be represented with an accounting model externally to OSeMOSYS (e.g. using WEAP) and soft-linked with OSeMOSYS. Several such applications have been developed. Many links between the energy, water and land systems are usually represented, e.g. water uses by power plants, diesel uses by agricultural activities, water uses by irrigation, and more. The links can be customised (and more can be added), depending on the scope of the analysis and the research or policy questions.

Key features of the CLEWs modelling framework

Fully open source, from solver to code, to pre- and post-processing codes, to interfaces, to data. Teaching material and documentation available on GitHub and on the Open University's OpenLearn platform (and others). Used in teaching, research and capacity development. Key in capacity development is that the tool is used, owned and kept by the trainees. Within 2 weeks, the trainees are guided into the independent creation of their own CLEWs model; then, the model is further developed with a COLLABORATION, rather than with a univocal or top-down approach.

Socioeconomic dimensions

CLEWs models have been linked to Input-Output models for the calculation of job loss and job creation effects, like done with energy system models.

Mitigation/adaptation measures and technologies

The technological representation can be entirely customised. Any mitigation option can be included, as long as it can be represented as a process with inputs, outputs, a transfer function and certain techno-economic characteristics. The use of adaptation vs -no-adaptation scenarios is common, to investigate the additional economic costs of investing in infrastructure that is not optimised for climate change.

Economic rationale and model solution

The model computes dynamically (year-by-year over the whole time domain) the investment and operation mix across the energy, land use and water sectors that minimizes the Net Present Cost of all sectors together while meeting commodity demands and other user-definedconstraints. The rationale is the same as for any energy system optimization model, but here land and water are included. The objective function and the constraints can be customized. Marginal costs of commodities can be extracted. Partial equilibrium behaviour can be reproduced. Myopic foresight can be introduced.

Key parameters

  • commodity demands for the energy sector, as well as the agricultural sector (crop and livestock products) and the water sector (water demands for public uses, industry, etc.)
  • techno-economic characteristics of energy supply technologies
  • commodity import prices (fuels, crops, water)
  • detailed costs of crop production per crop and per mechanization level
  • geospatial information of land covers and agricultural land uses in the base year
  • geospatial information on precipitation, evapotranspiration, surface water run off and groundwater recharge for the base year and for various climate scenarios

Key policies that can be addressed

Energy and climate policies (e.g. renewable targets, emission targets, emission penalties, targets of production of energy commodities or carriers such as hydrogen), agricultural policies (e.g. food security targets aiming at capping crop imports, incentives to domestic crop production, targets of increased water efficiency, targets of modernization of agricultural practices).

Implications for SDGs

It focuses on SDGs 2, 6, 7, 13 and 15.

Model presentation



Download slides in pdf

Recent use cases

Paper DOI Paper Title Key findings Long-term water–energy–food security and resources sustainability: a case study of Ethiopia by 2030 and 2050

The dominant hydropower energy capacity will be limited after 2040 and replaced by solar and other technologies. The target food production is constrained by limited land availability and productivity. Doubling of agricultural productivity is needed to effectively reduce the burden on land encroachment to sensitive ecosystems and reduce the dependency on imported food. It can be cautiously concluded that from resources supply–demand aspects, Ethiopia achieves middle-income country status by developing its untapped water and land resources.


Addressing water crisis and dying ecosystems in the Urmia and Hamoon basins (Iran) with an integrated planning perspective. Results of the CLEWs modelling show that there is not enough water to feed the restoration of the lakes as well as water demands for public uses and agriculture in the following years. The water deficit can be significantly aggravated by climatic change, which is particularly fast and intense in the two regions. The water deficit can be reduced by change in agricultural practices and import/export patterns with the rest of the country. The Impact of Climate Change on Crop Production in Uganda—An Integrated Systems Assessment with Water and Energy Implications Results, developed at a catchment level, indicate that on average there could be an 11% reduction and 8% increase in rain-fed crop production in the cumulatively driest and wettest climates, respectively. Furthermore, in the identified driest climate, the electricity required for pumping water is expected to increase by 12% on average compared to the base scenario.
Recent publications using the CLEWs model


Howells, M., Hermann, S., Welsch, M., Bazilian, M., Segerström, R., Alfstad, T., ... & Ramma, I. (2013). Integrated analysis of climate change, land-use, energy and water strategies. Nature Climate Change, 3(7), 621-626.

Ramos, E. P., Howells, M., Sridharan, V., Engström, R. E., Taliotis, C., Mentis, D., ... & Rogner, H. (2021). The climate, land, energy, and water systems (CLEWs) framework: a retrospective of activities and advances to 2019. Environmental Research Letters, 16(3), 033003.

Ramos, E. P., Sridharan, V., Alfstad, T., Niet, T., Shivakumar, A., Howells, M. I., ... & Gardumi, F. (2022). Climate, Land, Energy and Water systems interactions–From key concepts to model implementation with OSeMOSYS. Environmental Science & Policy, 136, 696-716.