Projected Global Fertilizers Consumption Datasets during 2020-2100 under SSP scenarios

Comprehensively predicting global fertilizer consumption in the future is essential for providing critical datasets in related fields such as earth system simulation, the fertilizer industry, and agricultural sciences. However, the absence of a reasonable fertilizer consumption prediction approach causes a scarcity in large-scale and long-time series fertilization datasets. Here, we produced a grid (5′×5′) global fertilizers (including N, P, and K fertilizers) consumption datasets from 2020 to 2100 under the shared socioeconomic pathway (SSP) scenarios, using an ensemble machine learning (EML) approach with six algorithms including multiple linear regression (MLR), decision trees (DT), auto-regressive integrated moving average (ARIMA), multi-layer perceptron (MLP), radial basis function (RBF) and random forests (RFs). Based on national fertilizer consumption (1961-2015), a long-time series of global historical fertilizer consumption datasets, the proposed EML approach was well-trained and the prediction models have been carefully validated with satisfactory accuracy. It indicates that the proposed EML approach provides a rational and reliable framework for fertilizer consumption prediction that stably outperforms the single algorithms with relatively high accuracy (Nash-Sutcliffe efficiency of 0.93, Kling-Gupta efficiency of 0.89, and mean absolute percentage error of 10.97 %). We found that global N and P fertilizer consumption may decrease from 2020 to 2100, while K fertilizer may buck the trend. N fertilizer consumption showed a declining trend of -1%, -17.13%, and -3.43% under the SSP1, SSP2, and SSP3 scenarios in 2100, respectively. For P fertilizer, those were -0.68%, -9.68%, and -2.03%. In contrast, global K fertilizer consumption may increase by 18.03%, 9.18%, and 6.74%, respectively. On average, N, P, and K fertilizer consumption is highest in China, and the lowest is in Kazakhstan. However, the hotspots of N fertilizer consumption may shift from China to Latin America and the Caribbean. We argue that the Projected Global Fertilizers Consumption Datasets are a valuable complement to currently available products. These datasets are expected not only to allow us to better understand the dynamics and distribution of global fertilizer consumption under different socioeconomic development paths in the future but also to provide support for relevant research. Such as but not limited to earth system simulation, the fertilizer industry, and agricultural sciences. Projected Global Fertilizers Consumption Datasets are stored in a zip package, that is PGFCD.rar. This package consists of 3 folders (N_fer, P_fer, and K_fer) and 1 file (Global Fertilizer Consumption (2020-2100).xlsx) once unzipped. Among them, Global Fertilizer Consumption (2020-2100).xlsx is the table of global fertilizers consumption projection for different countries and regions under SSP1, 2, and 3 scenarios 2020-2100. The folder of N_fer, P_fer, and K_fer contains the gridded global fertilizer consumption including the prediction of N, P, and K fertilizer under SSP1, 2, and 3 scenarios for every 10 years from 2020 to 2100, respectively. Each folder contains 27 GeoTIFF files, namely there are a total of 81 GeoTIFF files in the Gridded global fertilizers consumption projection datasets. Each GeoTIFF file is named as [Fertilizer type]_fer_con_[SSP scenario]_[year].tif. For example, K_fer_con_SSP1_2020.tif is the gridded global K fertilizer consumption projection for different countries and regions under the SSP1 scenario in 2020. The rest *.tif files can be recognized in the same way.