Understanding the climate projections

What are representative concentration pathways (RCPs) and how many are there?
Representative concentration pathways (RCPs) describe different future scenarios of greenhouse gas (GHG) concentrations (O’Neill et al., 2016). They specifically describe the GHG pathway to increased radiation levels by the end of the century. For example, RCP8.5 means an additional 8.5 W/m2 will reach the Earth’s surface by 2100. RCPs are based on various assumptions regarding, like population growth, economic and technological development as well as environmental protection, which are manifested in SSP scenarios (see below). Currently, there are seven RCPs, including RCP1.9, RCP2.6, RCP3.4, RCP4.5, RCP6.0, RCP7.0 and RCP8.5. We use RCP2.6 and RCP6.0 in earlier publications and RCP2.6 and RCP7.0 in more recent publications.

What are shared socio-economic pathways (SSPs) and how do they differ from RCPs?
For the latest IPCC Assessment Report 6, the climate modelling community has developed a new framework of scenarios which does not consider the RCPs as stand-alone scenarios, but rather as SSP-RCP combinations. Socio-economic pathways (SSPs) are narratives of future socioeconomic global changes (O’Neill et al., 2016). SSPs range from a sustainable scenario, which respects environmental boundaries, to a development, which relies on the exploitation of fossil fuels and the adoption of resource and energy intensive lifestyles. Currently, there are five SSPs, including SSP1 (Sustainability), SSP2 (Middle of the road), SSP3 (Regional rivalry), SSP4 (Inequality) and SSP5 (Fossil-Fueled Development). Each SSP can result in a range of possible GHG emissions and can, therefore, be combined with different RCP scenarios. However, the climate modelling community has defined a set of prioritized SSP-RCP combinations, out of which we use the combinations SSP1-RCP2.6 and SSP3-RCP7.0 (in climate risk analyses only).

What is the Intersectoral Impact Model Intercomparison Project (ISIMIP)?
ISIMIP is a community-driven climate-impact modelling initiative which is hosted at PIK and which aims at contributing to a quantitative and cross-sectoral synthesis of climate impacts, including associated uncertainties (Frieler et al., 2017). ISIMIP provides a common set of input data, such as climate data and socio-economic data to impact modelling groups and, therefore, produces consistent climate impact simulations and ensembles of impact simulations. Climate impacts are modelled for different sectors, for example water, agriculture, health, energy, forests and species. ISIMIP is currently moving from phase 2 to phase 3 of the project, with the different modelling groups updating their simulations with new input data, in line with the latest IPCC report.

What is the Coupled Model Intercomparison Project (CMIP)?
The Coupled Model Intercomparison Project (CMIP) is a collaborative initiative established by the Working Group on Coupled Modelling (WGCM) of the World Climate Research Programme (WCRP, 2022). Within CMIP, the scenario frameworks (e.g. RCPs and SSPs) are developed and common future emission and socio-economic input data is provided for the climate modelling groups. The result of CMIP is a consistent set of climate simulations ensuring to assessment of model uncertainties. CMIP is currently in its sixth phase (CMIP6) and includes more than 40 modelling groups.

How do ISIMIP and CMIP relate to each other?
In order to provide consistent input data for impact modelers, a selection of CMIP climate model projections (four CMIP5 models for ISIMIP2 and ten (five primary and five secondary) CMIP6 models for ISIMIP3) was bias-adjusted to observational data and downscaled to a common spatial grid of 0.5°.

Why did you use RCP7.0 as a worst case scenario and not RCP8.5?
We tend to use RCP7.0 as a worst case scenario, since it is a more realistic scenario, compared to RCP8.5. RCP8.5 has been used as a business-as-usual scenario, however, it has come to be viewed as increasingly unlikely and not plausible, given the current efforts to mitigate climate change, e.g. through a withdrawal from fossil fuels and other policies to limit CO2 emissions.

Which baseline periods do you use?
We use the time period from 1985 to 2014 (represented in the analysis by the central year 2000) as a baseline to represent current conditions. The only exception are air temperature projections in our climate risk profiles, for which we use the year 1876 as a baseline. We do this to show air temperature changes relative to pre-industrial levels, which allows us to put the result in context to the Paris Agreement.

Can you model the impact of climate change on other crops, for example, cotton?
Yes, we can model the impact of climate change on most crops. For the in-depth climate risk analyses, we select crops depending on their relevance in the country, their contribution to strengthening food security and their priorization in national strategies, among other factors. For this crop selection, we then set up context-specific crop models. For the much shorter climate risk profiles, we use the ISIMIP agricultural output, which is based on global crop models and which has a fixed selection of 12 crops.

Can you model the impact of climate change on livestock?
So far, our portfolio of impact models does not include a livestock model. However, we can approximate the impact of climate change on livestock by modelling the impact on grassland, which provides fodder for livestock.

Using the projections

The models are global. How are local data integrated, e.g. from local weather stations?
Local information is only included indirectly. The climate model simulations are bias-adjusted to observational datasets, which in turn include information from station-based datasets like CRU (Climatic Research Unit) temperature (Harris et al., 2020) or GPCC (Global Precipitation Climatology Centre) precipitation (Schneider et al., 2020).

Can the models project local climatic conditions?
Global climate models usually run on a spatial resolution of several hundreds of kilometers. Therefore, it is not possible to resolve very local processes. Regional climate models (not used in AGRICA) run on higher resolutions in order to capture these processes, but there has been little evidence that the increase in resolution reduces model uncertainty. While users of climate information often express a need for higher resolution, this would not necessarily increase the confidence in the information.

What is the maximum geographic resolution of the models?
While the climate models are run on scales of several hundreds of kilometers, the resulting simulations are downscaled to a resolution of 0.5 °, which corresponds to approximately 50 × 50 km near the equator. On this resolution, the impact models are then run. This means that for larger areas, it is easier to present a more differentiated picture of climate change and climate impacts, whereas for smaller areas, projections might be based on only a few grid cells, e.g. in the case of small island states.

Why are the model results uncertain?
Uncertainty in the climate projections stems from mainly three sources. First, natural variability of the climate system is largely unpredictable. Natural variability means climate fluctuations due to events like El Niño, volcanic eruptions or solar activity. Second, it is uncertain how future GHG concentrations are going to develop. To accommodate this uncertainty, RCP and SSP scenarios have been developed and we consider it vital to look at the range of these possible GHG pathway outcomes. The third major component to uncertainty in climate and climate impact assessments is to be found in the models and their calibrations themselves. While they represent a large number of the physical processes determining the climate, they are far from complete, as the climate system is vast and extremely complex. Many processes cannot be resolved due to resolution and computer power issues, other processes might not even be understood by science yet and, therefore, are not included. In order to assess this model uncertainty, we always consider an ensemble of different models under the assumption that different models can represent different processes well. In addition, models are set up and calibrated based on observational data. Uncertainties in observations contribute to uncertainties in model simulations and are hard to quantify, especially in regions with low density of station data.

If model projections are uncertain, how do they add value?
Making predictions about the future, we will always face a certain degree of uncertainty, regardless of possible improvements in modelling. Despite this uncertainty, models add value in that they help us to understand the earth’s complex climatic processes and to assess our climate’s response to increasing greenhouse gas concentrations. While not all aspects of future climate change might be clear for all regions, our large set of climate and climate impact indicators allows us to identify regional key issues and hotspots.

How can the projections be used for decision-making?
The projections can be used to inform various processes in adaptation policy and planning. Most importantly, they can serve to inform NDC and NAP processes, in addition to other national processes, strategies and plans. The projections can also be used for climate mainstreaming where climate is not yet a priority issue. While the main target group of our publications are policy makers at the national level, we always make sure to also translate the results for the end users, who are smallholder farmers, e.g. through farm radios or the production of videos. Further groups may also benefit from the results, including representatives from academia, the non-profit sector and the private sector.


Frieler, K., Lange, S., Piontek, F., Reyer, C. P. O., Schewe, J., Warszawski, L., Zhao, F., Chini, L., Denvil, S., Emanuel, K., Geiger, T., Halladay, K., Hurtt, G., Mengel, M., Murakami, D., Ostberg, S., Popp, A., Riva, R., Stevanovic, M., … Yamagata, Y. (2017). Assessing the Impacts of 1.5°C Global Warming – Simulation Protocol of the Inter-Sectoral Impact Model Intercomparison Project (ISIMIP2b). Geoscientific Model Development, 10, 4321–4345. https://doi.org/10.5194/gmd-10-4321-2017

Harris, I., Osborn, T. J., Jones, P., & Lister, D. (2020). Version 4 of the CRU TS monthly high-resolution gridded multivariate climate dataset. Scientific Data, 7(1), 1–18. https://doi.org/10.1038/s41597-020-0453-3

O’Neill, B. C., Tebaldi, C., Van Vuuren, D. P., Eyring, V., Friedlingstein, P., Hurtt, G., Knutti, R., Kriegler, E., Lamarque, J. F., Lowe, J., Meehl, G. A., Moss, R., Riahi, K., & Sanderson, B. M. (2016). The Scenario Model Intercomparison Project (ScenarioMIP) for CMIP6. Geoscientific Model Development, 9(9), 3461–3482. https://doi.org/10.5194/gmd-9-3461-2016

Schneider, U., Becker, A., Finger, P., Rustemeier, E., & Ziese, M. (2020). GPCC Monitoring Product: Near Real-Time Monthly Land-Surface Precipitation from Rain-Gauges based on SYNOP and CLIMAT data. https://doi.org/http://dx.doi.org/10.5676/DWD_GPCC/MP_M_V2020_100

WCRP. (2022). WCRP Coupled Model Intercomparison Project (CMIP). https://www.wcrp-climate.org/wgcm-cmip