The annual TRENDY MIP was started in 2010 in order to compare different land models and their ability to model the sources and sinks of carbon in the land (i.e., to/from the soil and the vegetation from/to the atmosphere). TRENDY provides model-mean and model-spread estimates of the uptake of carbon by the land throughout the globe for the Global Carbon Project, resulting in an annual publication about the Global Carbon Budget. The Global Carbon Budget attempts to balance the carbon budget for the entire globe, including the anthropogenic carbon emissions (whether from burning fossil fuels or from changes in land use) and the changing carbon reservoirs in the ocean, atmosphere, and the land. As far as I know, TRENDY is not a typical acronym, where each letter corresponds to a word, but instead, it refers to ‘trends’ in the carbon stores and fluxes with time.
I arrived at the University of Reading from my former home in Berlin in July 2017 to work as a Land Surface Processes Computational Scientist here. My involvement in the Global Carbon Project’s TRENDY MIP can be traced back to January 2018 when Tristan Quaife (University of Reading) asked me to convert data for the CALIBRE agroforestry project. He wanted me to take the forcing/driving weather data from ISIMIP up to the year 2100, which was in NETCDF format, and convert it to an ASCII/text format used by the Sheffield Dynamic Global Vegetation Model (SDGVM; Woodward, Smith, and Emanuel, 1995; Woodward and Lomas, 2004; Walker et al., 2017). The idea was to then run SDGVM with this weather data and with a variety of agroforestry reforestation scenarios in the UK in order to see if there was an optimal agroforestry strategy. I was not involved in running the SDGVM model further for the CALIBRE agroforestry project, but my work on converting the driving weather data for SDGVM did catch the eyes of both Tristan and his colleague, Anthony Walker. Anthony has been a developer of SDGVM for a number of years, and he is based at Oak Ridge National Laboratory in Tennessee, in the USA.
Anthony and Tristan subsequently asked me in March 2019 if I could use SDGVM to participate in a MIP for comparing land models for the European Drought in 2018, which has been led by Ana Bastos. Ana was then at Ludwig Maximilians University in Munich, and she is now at the Max Planck Institute for Biogeochemistry in Jena, Germany. For this project on the 2018 European Drought, I was able, firstly, to adapt Anthony’s R scripts to convert the ERA5 driving data (from the NETCDF format to SDGVM text format) over the European region for 1979-2018 that Ana had provided; secondly, to run SDGVM with this converted data; and thirdly, to produce the daily and monthly NETCDF output files (converting from SDGVM text format to NETCDF format) for SDGVM. I then delivered those NETCDF output files from SDGVM to Ana, which she analyzed for the 2018 European Drought MIP. She has subsequently published three first-author journal papers, including myself as the unique SDGVM co-author, in Bastos et al. (2020a-b, 2021). One rule for the 2018 European Drought MIP was that only one member of each model’s team could be a co-author. This rule was passed down to the 2018 European Drought MIP from the TRENDY MIP’s set of rules. Otherwise, surely Anthony and Tristan would have been co-authors, as well.
From what I understand, some years earlier, Tristan had approached the TRENDY MIP organizers, suggesting that Anthony, as a SDGVM expert, could run SDGVM for the TRENDY MIP, and hence, SDGVM could further complement the other DGVMs that were already part of the TRENDY MIP. Anthony was able to get involved in the TRENDY MIP with SDGVM, and the first year that he was a co-author on the subsequent annual Global Carbon Budget paper was in 2016 (Le Quére et al. 2016). Anthony also ran SDGVM for the TRENDY MIPs, which were published in 2017, 2018, and 2020. Due to my work on the 2018 European Drought with SDGVM, Anthony and Tristan asked me to run SDGVM for the TRENDY MIP, which was published as part of the Global Carbon Budget in 2019 (Friedlingstein et al. 2019). This involved the conversion of CRU weather input data (Harris et al. 2014) from NETCDF format to SDGVM text format and the landcover data (Land-Use Harmonization based upon LUH2 v2h from 1700-1950, and based upon HYDE from 1951-2018) from NETCDF to the SDGVM text format; running the SDGVM under several different TRENDY experimental protocols; and converting the SDGVM text-formatted output files to NETCDF format. This sounds straightforward, but it also involves checking for mistakes and ensuring that the output data is reasonable and makes sense. Overall, this part of the process started for me in mid-July 2019, and I delivered the first SDGVM runs to the TRENDY team in mid-August 2019, and the publication of the GCB paper was in December 2019. The Global Carbon Budget paper in 2019 already has almost 800 citations! Anthony did the SDGVM runs for TRENDY in 2020, and I was asked to do the SDGVM runs for TRENDY this year (2021), wherein I made my first delivery this year in mid-September. The 2021 Global Carbon Budget publication (Friedlingstein et al. 2021) should be submitted in a week or so from the time of publication of this blog. So stay tuned!