Data k článku. Climate-induced Severe Water Scarcity Events as Harbingers of Global Grain Price

Fig. S1. Weights of the grids used in this study. a-c) Wheat, maize and rice grids: Grids where the given crop is grown, based on Ramankutty et al. (55), with the weights based on the overall proportion of arable land; d-f) the weights were assigned based on the share of the wheat-, maize- and rice-growing area in the grid in relation to the crop-growing area of the top ten exporters; g-i) indication of the main producing areas and potentially irrigated area of the given crop. Fig. S2. Time series of severe water scarcity (SWS) for the entire wheat-growing area (a, c) and area of wheat cultivation of the top 10 wheat exporters (b-d) for the baseline climate and projections based on the CMIP5 (a, b) and CMIP6 (c, d) GCM models. Fig. S3. Matrixes of all tested models SWS over arable land (a) and SWS over wheat producing area of 10 largest wheat exporters (b) vs. wheat price index (WPI). The yellow marked combinations were used for the final model (the ensemble of 5 best performing models called Final model). At (c) the performance of other tested approaches (ARIMA and SPEI based models) with the Final model is shown The annually resolved reported wheat prices and their estimates by 5 different approaches for (2000-2021) and the model projections till 2100 are shown for comparison at Fig (d). Fig. S4. a) Reported farm price index (FPI) and estimated wheat price index (WPI), which was modeled based on the final model ensemble for the 1951-2023 period with the RMSE of the model estimates provided for 3 periods and the entire 73 years. B) The same figure but with the colors marking the total arable land area affected by SWS during the three-year window. Note: years 2022 and 2023 were not used in development of the wheat price index estimate model. Fig. S5. Same as Fig. 3 but based on the CMIP5 ensemble of GCM models. Fig. S6. Comparison of the wheat price index estimates based on the CMIP5 and CMIP6 GCM model ensembles. a) Difference between the wheat price index (WPI) values estimated by CMIP6 and CMIP5; b-c) relationship between the global mean temperature change and WPI. Fig. S7. Comparison of the literature on the wheat price index. a) WPI values estimated by climate change scenarios for CMIP5 and CMIP6 in this paper, Global Biosphere Management Model (GLOBIOM) data vs. the literature between 2000 and 2050. b) Percentage of change in the WPI related to the baseline of this study and the literature for the different levels of global warming (change of global temperature related to mean 1951 - 1980). Fig. S8. Top 10 key player countries in the wheat, maize and rice market between 2000 and 2019 from FAOSTAT. a) Top 10 cumulative exporter countries: sum of total export quantities between 2000 and 2019. b) Top 10 exporter countries: mean of the total export quantities between 2000 and 2019. c) Top 10 net exporter countries: mean of the total export minus total import quantities between 2000 and 2019. Table S1. Definitions of severe water scarcity (SWS) used. All conditions must be fulfilled for the evaluated climate grid cell to be categorized as experiencing severe or extreme water scarcity during a given year. The sensitive period (SP) was defined as the four months prior to the harvest of a particular crop. The standardized precipitation evapotranspiration index (SPEI) is employed as it encodes the imbalance between precipitation and potential evapotranspiration. This imbalance captures the duration and severity of a drought event from hydroclimatic variables at large (i.e. climate grid) scales. Table S2. Overview of the best performing linear models for estimating the wheat price index (WPI) based on the area affected by severe water scarcity. The aggregated values indicate the combined performance of the models as the ensemble is expanded by adding additional models to Model_1. The bootstrapping results are part of Fig. 2b. Table S3. List of the CMIP5 global circulation models (GCMs) used in this study with brief descriptions. The historical run (i.e., 1850‒2005) and the runs of three future emissions scenarios (i.e., RCP2.6, RCP4.5 and RCP8.5; 2006‒2100) are used from each model. “*” indicates that the simulations under all three future RCPs are available. Otherwise, only RCP4.5 and RCP8.5 are available. The first ensemble run is used if a model is associated with multiple ensemble runs. Table S4. Overview of the CMIP6 global circulation models (GCMs) used in this study (with brief descriptions). The historical run (i.e., 1850‒2014) and the runs of three future emissions scenarios (i.e., SSP1-2.6, SSP2-4.5 and SSP5-8.5; 2015‒2100) are used from each model. “*” indicates that the simulations under all three future SSPs are available. Otherwise, only SSP2-4.5 and SSP5-8.5 are available. The first ensemble run is used if a model is associated with multiple ensemble runs. Table S5. The sources of export prices used for calculation of crops specific price indices (WPI, MZI and RPI) by the International Grain Council. The export prices are provided for the nearest available shipment and are based on both official and trade source. The FOB (Free On Board) price is the price of goods at the frontier of the exporting country or price of a service provided to a non-resident. It includes the values of the goods or services at the basic price, the transport and distribution services up to the frontier, the taxes minus the subsidies.