This opinion article was written by Scientific Analysis Manager Lillian Gordon.

Climate is changing at different rates all over the globe. Some places may experience an increase in extreme storms and precipitation. Others may see skyrocketing temperatures and prolonged droughts.  

No matter the specifics, these shifting environmental factors can have major impacts on human migration, disease transmission, agricultural production, and freshwater supplies, making it important for decision makers at all levels to be aware of the changes they may face in the future.  

How can decision makers learn more about the impact of climate change? 

To increase this awareness, some scientific institutions have created interactive climate dashboards that provide a visual representation of past and future climate variables over varying geographic extents. These dashboards are dense with information, which can deter first time climate interpreters from getting their feet wet. This article is meant to provide the foundational questions that early climate model interpreters should use when first assessing any climate dashboard.  

Climate dashboards can vary widely in their construction and therefore need thoughtful consideration when assessing data in the displays. When looking at a dashboard for the first time, there are four aspects you should always check: datasets, climate scenarios, date ranges, and variables.  

What makes a good climate dashboard? 

Data Sources 

When looking into climate forecasting, the most important detail is where the data is sourced. A dashboard may hold many sources of data, but a simple first check is to look at the sources of the historical and forecast datasets.  

In a historical dataset, the data can be either observed (e.g., ERA5 that uses part observed data and part modeled data to create a complete record of past climate or a modeled historical dataset (e.g., CMIP6 that uses only modeled data). The historical dataset is important because it provides the baseline with which to compare future forecasts. 

In most forecast data, the outputs are derived from a suite of models, such as CMIP5 or CMIP6 and are narrowed further by the dashboard host. Therefore, it is worth noting which models are a part of the suite because differences in models can cause differences in model outputs. It should also be noted that the downscaling technique for both historical and forecast data is important because one technique over another could change the model outputs.  

Climate Scenarios 

After deducing the source of the forecast data, it is important to find out the climate scenarios being presented for the future. Climate scenarios provide possible futures that climate scientists can test their models on. There are two components for these scenarios: the Representative Concentration Pathways (RCPs) and the Shared Socioeconomic Pathways (SSPs).  

RCPs such as RCP 2.7 and 8.5 strictly quantify the concentration of greenhouse gases and the radiative forcing due to increases in climate change pollution. Essentially, the higher the RCP, the more radiative forcing you will have due to an increase in GHG pollution.  

SSPs, on the other hand, consist of five different narratives that describe social and economic futures through combined qualitative narratives and quantitative inputs on population, GDP, and urbanization. The narratives range from 1 to 5 with SSP1 being a sustainability focused future and SSP5 being a fossil fuel development future.  

It is important to be aware of which climate scenario is being used due to these differences in narratives. Outputs from CMIP5 will only consider radiative forcing whereas the outputs from CMIP6 will consider both radiative forcing and additional aspects of human influence.   

Date Ranges 

Next, it is important to consider the past and future date ranges that are presented. Historic date ranges are important because this acts as the baseline that will be compared to all future forecast predictions. Therefore, it is important to know if the historic range is from the past 10 years or past 50 years.  

This all depends on the question you are asking. Forecast date ranges can be more variable in how they are presented, either as a range of dates or a single year, but even single years are computed by an average year. However, a rule of thumb with global climate models is that projections between 1 to 15 years in the future tend to be less accurate since there are many influences of current weather on these timescales. Projections 20+ years out tend to yield better projections.  

Simple and Complex Variables 

Now that the inner workings of the dashboard have been defined, the variables presented in the dashboard can be understood with better confidence and clarity. Variables tend to come in two categories: simple and complex.  

Simple variables are those that do not change based on geographic location. For example, maximum temperature will always be the maximum temperature no matter the region. Complex variables, on the other hand, do change with reference to their geographic location and need closer questioning when being observed. This includes variable like “extreme heat” meaning extreme heat thresholds are going to be different, for instance, in Arizona versus Minnesota. Knowing how these complex values are calculated better inform the user to the changes that will be existing in their region of interest.  

Dashboards can be complicated places to navigate but when the four items above are understood and put into practice, decision makers can be better equipped to deal with the changing climate.