CMIP5 Data Analysis Tool - Help

This webpage contains details and instructions for the web-based interactive CMIP5 Data Analysis Tool.

To go to the user input page please follow the link below.


Experiment vs Variable and Model vs Ensemble

The first table allows for the selection of a variable (column headers) and an experiment (row headers). In case the headers are not clear then hovering over the header names will show some details. Depending on the variable-experiment selection the available check boxes in the second table will change indicating which models and ensemble members are available.

The second table allows fo the selection of ensemble members (column headers) and models (row headers). The Select All checkbox can be used to select al models and ensemble members with one mouse click. Selecting whole columns or rows can be done by clicking on the header labels.


Time Period

Select the time period you are intersted in here. Depending on which experiment was selected in the Experiment vs Variable and Model vs Ensemble section above some years will be disabled. For instance, if historical runs are selected only years between 1850 and 2005 can be selected. Similar, if RCP* runs are selected then only years from 2006 onwards can be selected.


Month/Season

Select All for either Month or Season to include all 12 months for the calculation. Choosing the period Jan to Dec will have the same effect. Use the dropdown menues to select a specific month or a standard season. In addition, a an arbitrary period between two months can be selected.

Note that when a period is selected that spans two years a special caveat has to be considered. For exampel, if the period Nov-Mar was selected for the years 2000 to 2010 then in the first year of the timeseries (2000) the months Jan, Feb and Mar will be selected in addtion to Nov and Dec. Similar in the last year of the timeseries (2010) the months Nov and Dec will be selected in addition to Jan, Feb and Mar.


Plot

Select Absolute value to calculate the absolute value of the selected field. If more than one model run was selected then the Multi Ensemble/Model Options will be enabled which allows for the calculation of Mean, Minimum, Maximum, Variance, Standard Deviation or Percentile.If more than one model and more than one ensemble was selected then the ensemble mean (or the mean of the ensemble anomalies) for each model is calculated first and the Mean, Minimum, Maximum, Variance or Standard Deviation is then calculated based on the model ensemble (anomaly) means.

If Percentile is selected then a percentile value must be entered into the corresponding text field. The percentile value must be between 1 and 99.

Select Anomaly with respect to YYYY to YYYY historical climatology and choose a reference time period if you want to create anomaly plots. The reference field will be from the historical experiment and the reference time period therefore can be selected only between 1850 and 2009. If an RCP*.5 experiment is selected make sure that a historical run exists for this model, ensemble member and variable, otherwise, the calculations will crash. The reference field will be calculated respecting the Month or Season selected in the Month/Season section.

If Anomaly with respect to YYYY to YYYY historical climatology and either All Models or All Ensemble Members is selected then the anomaly is calculated for each ensemble member first and then the Mean, Minimum, Maximum, Variance, Standard Deviation or Percentile of the ensemble anomalies is calculated next. If Anomaly with respect to YYYY to YYYY historical climatology and both All Models and All Ensemble Members is selected then mean of the ensemble anomalies is calculated first and then the Mean, Minimum, Maximum, Variance, Standard Deviation or Percentile is calculated based on the ensemble anomaly means.

If only one model was selected then the model's resolution will be kept. If more than one model was selected then they need to be remapped to the same resolution. By default data will be remapped to a 1 by 1 detree regular grid using nearest neighbor sampling. Choose a different interpolation method and/or resolution if required.


Plot Type

This section allows the user to select how the output field is supposed to be calculated. Selecting Show time average as will calculate the mean over the temporal domain whereas selecting Show time series for will calculate a time series by calculating a field average for the selecting domain at each time step. For time average plots the user has the option of producing a map, a zonal mean or a meridional mean.

For time series plots the user has the option to produce area mean, zonal mean or meridional mean. For time series plots annual means are calculated.


Spatial Domain

The user can select a pre-defined model domain by using the radio buttons. Alternatively, set the model domain manually by selecting the Custom radio button and editing the Lowest Lat, Highest Lat, Lowest Lon and Highest Lon fields. The Latitude range goes from -90 to 90 and the longitude range goes from -180 to 180.


Contour Plot Options

Most of the options in this section are self explanatory.

Plot Type

  • Filled Contours - Plots the field as filled contours.
  • Blocks - Plots the field as boxes. Box size depends on model resolution.
  • Isolines - Plots the field as black isolines with labels on them.
  • Zero Line - If checked, draws a dotted line at the zero contour level on anomaly map plots.
Contour Levels
  • Min - Lowest contour level.
  • Max - Highest contour level.
  • Interval - Step between contour levels.
  • Max must be larger than Min. If no contour level details are provided then the minimum and maximum of the field are used the interval is calculated automatically by the IDL routine.

Colour Scale
  • # - Select colour map to be used for plotting.
  • Reverse Colours - Reverses the selected colour scale.
  • Colour map names can be checked here.

Colour Bar
  • Right - Plots colour bar to the right of the plot.
  • Below - Plots colour bar below the plot.
  • No Colour Bar - Plots no colour bar.
Colour Bar Options
  • Upper Open End - Plots a colour bar with an upper open end.
  • Lower Open End - Plots a colour bar with an lower open end.
  • n Decimals - Defines the number of decimals to be used in the colour bar.
Map Features
  • Countries - Plots country outlines.
  • Hi Res - Draws the coastline of continents in high resolution.
  • Lakes - Draw lakes such as the Great Lakes in the US or the Caspian Sea.
  • Isotropic - Scales longitude and latitude equally.
  • Turning the Isotropic option on and off may be useful in some cases to adjust the plot size on the page.

Map Projection
  • Cylindrical - Use Cylindrical projection.
  • Robinson - Use Robinson projection.
  • Mollweide - Use Mollweide projection.

Global Plot Options

Grid Lines

  • No - No grid lines are drawn.
  • Full - Horizontal and vertical grid lines are drawn.
  • Horizontal - Horizontal grid lines are drawn.
  • Vertical - Vertical grid lines are drawn.
Page Format
  • Landscape - Landscape is used as page format (default).
  • Portrait - Portrait is used as page format.
  • Switching between landscape and portrait may be useful in some cases to adjust the plot size on the page.

Known Issues

  • When both All Models and Show time series for are select the calculations may crash because the number of records (timesteps) varies between models.
  • The Lakes checkbox does not work because of a bug in the IDL function.
  • The n Decimals entry is applied to the colour bar only.
  • There is currently no option to control the number of decimals on the x and y axis.
  • The Grid Lines option may behave oddly due to bugs in IDL function.


Contact

Sebastian Engelstaedter
Climate Research Lab (CRL)
Oxford University Centre for the Environment (OUCE)
South Parks Road, Oxford OX1 3QY, UK
phone: +44 (0)1865 275867
email: sebastian.engelstaedter(at)ouce.ox.ac.uk

Please report any bugs !!!


Acknowlegements

The calculations are done using CDO. The plots are created using IDL. Special thanks go to A. Heaps (NCAS) for developing great IDL plotting routines (Using IDL in Meteorology).