GCCV Tutorial

Using the GCCV application:

The Global Climate Change Viewer (GCCV) is used to visualize future temperature and precipitation changes simulated by global climate models in the Coupled Model Intercomparison Project Phase 5 (CMIP5).  The application allows the user to visualize projected climate change (temperature and precipitation) for each country, for all available models and all Representative Concentration Pathways (RCP) emission scenarios (2.6, 4.5, 6.0 and 8.5 [more info]).  Data from the experiments are binned into 25-year climatologies that span the 21st century.  The GCCV also provides access to plots and quantile breakdowns of monthly temperature and precipitation from 1850-2100.  In addition, the application includes access to the currently available model simulations from the Last Glacial Maximum (LGM; 21 ka) and mid-Holocene (6 ka), which are part of the Paleoclimate Modelling Intercomparison Project phase 3 (PMIP3).  The primary data on which the GCCV is based can be downloaded in NetCDF format using the Earth System Grid Federation (ESGF) data portal.  The GCCV application only includes models with unrestricted access, which is a subset of all models available through the ESGF data portal.

The main window:

The main window of the GCCV (shown above) displays maps of either temperature or precipitation anomalies (the change between a future experiment and the present day or control simulation).  The maps emphasize the spatial differences of climate change globally and regionally.  The climatology plot (lower left) compares monthly averages and standard deviations (vertical bars), which are a measure of variability.  The histogram (lower right) displays the distribution of anomalies for all model simulations included for a particular experiment and is a quick way to visualize the spread of the simulated climate change anomalies for a given country.

Navigating options in main window:

The dropdown menus across the top of the application (outlined in red in the above screenshot) are used to select either annual or monthly means, the average of all model simulations or a specific model simulation for either temperature or precipitation.  The country of interest can be changed either by selecting it from the dropdown list of countries in the control bar or by clicking on the map.

The project (CMIP5/PMIP3), experiment and time period displayed in the anomaly map and charts can be changed in the Select Datasets window, which is found in the upper right corner of the application.  Clicking on the button opens a menu (in the figure below) that allows the user to select the project (CMIP5 or PMIP3) the experiment (RCP, LGM or mid-Holocene), and the time period of interest.

The application currently includes data for the following CMIP5 experiments:

CMIP5 # Models 1980-2004 2025-2049 2050-2074 2071-2095
Historical variable* X
RCP 2.6 19 X X X
RCP 4.5 23 X X X
RCP 6.0 15 X X X
RCP 8.5 26 X X X

* The number of Historical models used varies with each future RCP scenario.  For example, the 26 RCP 8.5 model simulations will be paired with 26 Historical simulations to calculate the climate change anomalies.

Some modeling centers have not submitted data for all RCP experiments so the RCP scenarios have varying numbers of models.  The GCCV includes the currently available data (as of Dec. 1, 2012) from the modeling groups that have made their data available without restrictions.

Selecting PMIP3 in the project menu provides access to the paleoclimate experiments that include pre-industrial, mid-Holocene (6 ka) and Last Glacial Maximum (21 ka) simulations (shown below).  At this time, fewer paleoclimate experiments have been submitted than CMIP5 experiments.

Summary climatology plots:

The climatology plot in the bottom left of the application (shown above) compares the control simulations (1980-2004, blue line) with observations (CRU TS3.10 and TS3.10.01 for 1980-2004, black line) and the future changes (here, 2050-2074, red line) for the selected country, model(s) and selected RCP or PMIP3 experiment.  The vertical error bars indicate the standard deviation of each month for the respective time periods.  In the example above, the temperature of the control simulations over the United States compares well with the observations during the summer months but displays a cold bias during the winter.  A 3-4 °C warming throughout the annual cycle is simulated for the 2050-2074 period.  Hovering the cursor over a particular month displays values for the mean and standard deviation of the observations, control and future simulations.

Clicking the Summary Table button located in the bottom right of the climatology plot provides a table of additional quantitative information about the simulations (shown above).  The columns in the table list the model, the annual root mean square error (RMSE) between the simulated and observed data, the annual or monthly mean and standard deviations for the control run (1980-2004), the annual or monthly bias between simulated and observed values and the annual or monthly future change.

The annual RMSE is a measure of how well the simulated seasonal cycle matches that of the observations.  The annual or monthly bias is the difference between the model control (present) simulation and the observations (calculated as control – observation).  Clicking on the header of a column sorts the column in either ascending or descending order.  Clicking on a row in the table changes the selected models shown on the map and climatology chart.  Clicking on the Plot button opens a new browser window that displays monthly time series of temperature and precipitation from 1850 through 2100 and an accompanying quantile data table (shown below).

In the table, data for the simulated periods are ranked into quantiles, which are useful in understanding projected changes in extreme and central values for a given country, experiment and model combination.  The time series and data table figures are not available for the paleoclimate PMIP3 simulations.

For the paleoclimate data in the GCCV, the black Observations line in the climatology chart, Annual RMSE and Bias columns in the Summary Table are all relative to modern observations from CRU TS3.10, whereas piControl is the pre-industrial simulation produced by the PMIP3 models.

The histogram plot in the bottom right of the application (shown below) portrays the distribution of climate change simulated by all the models for the selected country, experiment and time period.

Bins for temperature or precipitation change are indicated on the horizontal axis and the percent of models falling within each bin are indicated on the vertical axis.  The histogram gives a sense of the range and distribution of climate change simulated by the models.  Hovering the cursor over the histogram bars produces a window that summarizes the distribution.  The information in the above example indicates that the mean temperature change for all the models is 3.6 °C; ~15% (4 out of the 26 total models) fall within a range of change of 2.5-3.0 °C.  The names of the models in the selected bin are also given.  Clicking on a histogram bar multiple times cycles through all the models in that bin and changes the anomaly map and climatology plot to the selected model.

Modeling center information and links to data:

Information about the models and their respective supporting centers and access to the primary data are accessed from the Model Info button in the upper left of the application.  The Data button will open the ESGF data portal where the primary model data can be downloaded in NetCDF format.

Start using the GCCV!