Climpact2 and PRECIS WMO Workshop on Enhancing Climate Indices for Sector-specific Applications in the South Asia region Indian Institute of Tropical Meteorology Pune, India, 3-7 October 2016 David Hein-Griggs Met Office Hadley Centre, UK
Crown copyright Met Office Resolution is important (example)
Crown copyright Met Office Resolution is important (example)
Weather vs. Climate Climate is what we expect, weather is what we get (1887) Climate = Average weather and its variability over a period of time, ranging from months to millions of years. WMO Quantifies climate over a 30-year average period. Observations create the 30 year climate baseline, which new observations and climatic trends are measured against. Climate refers to the state of the climate system as a whole, including a statistical description of its variations.
What is a Regional Climate Model? An RCM is a mathematical model of the atmosphere and land surface, with a representation of the ocean surface High resolution: Produces data in grid cells 50km in size Spans a limited area (region) of the globe An RCM contains representations of many of the important physical processes within the climate system Cloud Radiation Rainfall Atmospheric aerosols Soil hydrology Etc.
Advantage: RCM over GCM Potential advantage of RCM vs GCM is related to the representation of spatial scale smaller than 300km and/or temporal scale smaller than 30 minutes these are absent in GCM Di Luca et. al (Climate Dynamics, 2012)
RCMs describe daily extremes more accurately Frequency of winter days over the Alps with different daily rainfall thresholds. RCM and Observations aggregated at GCM scale
RCMs are able to resolve intense mesoscale systems A tropical cyclone is evident in the RCM (right) but not in the GCM
A chain of separate models Global Climate Model Regional Climate Model Impacts Models
ClimPACT2 An R software package for calculating climate extremes indices, authored by a team led by Assoc. Prof. Lisa Alexander and Nicholas Herold at the University of New South Wales. https://github.com/arccss-extremes/climpact2
ClimPACT2 in gridded mode
Installation tutorials Climpact2 analysing climate model data works in Linux only I nstall R in (Ubuntu) Linux https://www.youtube.com/watch?v=t_0lbtxvs5s Run Climpact2 GUI in Linux https://www.youtube.com/watch?v=1sscqsrmczi
Proof of concept: ClimPACT2 analysing PRECIS output data I ran two PRECIS experiments for the case study over a 25km Eastern Caribbean region. One experiment downscaled 1980-1990 from the HadGEM2-ES GCM (a CMIP5 model) using historical GHG values and a second 2080-2090 from HadGEM2-ES RCP 8.5.
Proof of Concept: Climpact2 analysing PRECIS output data The raw output data from the experiments has to be processed to be Climpact2 ready. That means regridding, changing the units (to Celsius / millimetres per day), conversion to NetCDF data format, and more. To do this, NCO tools and CF-Python are needed. http://nco.sourceforge.net/ http://cfpython.bitbucket.org/
Proof of Concept: Climpact2 analysing PRECIS output data When the PRECIS data was in the right format, I edited a file in the climpact2-master directory: climpact2.ncdf.wrapper.r climpact2.ncdf.thresholds.wrapper.r I made changes to this file to the names of my variables (e.g. Air_temperature instead of Tmax) and to the input file names.
Proof of Concept: Climpact2 analysing PRECIS output data I then ran: Rscript climpact2.ncdf.wrapper.r inputting the baseline (historical period) data Next I ran: Rscript climpact2.ncdf.thresholds.wrapper.r and wrote out a quantiles file (percentiles of certain variables) Finally Rscript cllimpact2.ncdf.wrapper.r inputting the future projections data and the quantiles file. Climpact2 ran and produced output NetCDF files.
Next Steps If you want to use Climpact2 to analyse regional climate data, you can either generate data yourself with an RCM (including PRECIS) or download data from online repositories, e.g. the Coordinated Regional Downscaling Experiments project: http://www.cordex.org There are a number of projects which have run RCMs over South Asia. Some of them are willing to share their data.
The CMIP5 Global Climate Model CNRM-CM5 (Meteo France) has been downscaled by PRECIS over the DECCMA project South Asia domain Case Study : Climpact2 analysing PRECIS output data
Horizontal Resolution: 25km*25km Baseline: 1961-1990 Future: 2070-2099 Greenhouse Gas pathway: RCP8.5 Case Study : Climpact2 analysing PRECIS output data
Panoply Panoply is a netcdf, HDF and GRIB data viewer written by NASA-GISS. It uses a mouse drive, easy to use interface Panoply is written in Java and needs Java to be installed. Panoply runs on Windows, MacOS or Linux Panoply plots data from NetCDF, HDF and GRIB data sets. You can: Slice and plot specific latitude-longitude, latitudevertical, or time-latitude arrays from larger multidimensional variables. Overlay continent outlines or masks on lon-lat plots Change the scale and colour table Save plots to disk GIF, JPEG, PNG or TIFF bitmap images or as PDF or PostScript graphics files.
Panoply The main window
A sample plot of precipitati on using Panoply
Case Study : Climpact2 analysing PRECIS output data (Panoply live demonstration)
Interpreting Climpact2 Activity For the rest of the session you will be doing group work looking at climpact2 output data for: Drought Rain HEAT
Questions?
Optional Ending Activity follows
Station data vs.. Gridded data David and Victoria Beckham have just purchased a 27 million pound country house in England. David wants to know if he needs to adapt his house for climate change, so he runs PRECIS over the UK. When the models finish he then extracts the grid box his house is in and starts performing analysis on the data for that grid box. What s wrong with his method? Crown copyright Met Office
Station data vs. Gridded data Compare like with like Data only have skill at spatial scales resolved by their grids The grid box data values are an area average for the full area. They are not point data and therefore not directly comparable with single point time series. Crown copyright Met Office