Stories 2 - MAgPIE stories #2

Reactive nitrogen is essential for agricultural production, but the overuse of synthetic nitrogen fertilizer, especially in China, has imposed challenges to food security and the environment. Enhancement in soil nitrogen uptake efficiency (SNUpE) has been evidenced as one of the most effective means of reducing fertilizer use. However, a wide range of SNUpE and assessing their impacts on environmental and economic outcomes warrant exploration, as the future socioeconomic development trajectory remains unclear. We use MAgPIE alongside an empirical method (Difference-in-differences, DID) to comprehensively analyze the impacts of improving SNUpE on the environment and food security. Our model results show that enhancing SNUpE can largely reduce nitrogen fertilizer use and abate nitrogen pollution in the short and long term. Additionally, our findings reveal that the enhancement of SNUpE exhibits substantial co-benefits for the environment and food security.

This presentation gives an insight into one of the madrat-R-libraries of MAgPIE’s preprocessing. The goal of this hydro-economic model is to bridge modeling scales, bringing together spatially-explicit hydrological processes with economic information and preparing the data for usage in MAgPIE. It highlights the role of pre- and post-processing of data for global land-use modeling and discusses the trade-offs in adding complexity to MAgPIE’s core versus simplification by “outsourcing” code to R.

In this session, I present ongoing work focused on spatial validation of cropping patterns in India within the MAgPIE framework. MAgPIE currently derives cropping patterns based on yield and biophysical constraints taken from the LPJmL model. However, actual cropping patterns in India deviate significantly from these modeled allocations due to socio-economic, policy, and infrastructural factors. To improve the realism and regional relevance of MAgPIE outputs, I am incorporating state-wise historical data on crop area, production, and yield for major food grains for years 1990-2024, sourced from government databases. To embed this, I am building on the mrfable package, where I add new ‘read’, ‘correct’, and ‘calc’ functions to process and integrate the data and support validation in MAgPIE. The updated package aims to provide a robust basis for aligning modeled outputs with observed agricultural practices in India, enabling more accurate scenario analysis and policy insights.

This MAgPIE story describes recent advancements in the development of the MAgPIE-Brazil model, exploring challenges and solutions for Brazil. Starting with the motivation for this effort, we compare MAgPIE outputs with Brazilian national datasets, showing that while the model does not precisely match historical and spatial values of some land use classes, the trend of MAgPIE’s evolution during matches the historical period and official datasets. To improve how the model captures the dynamics of Brazilian land use patterns, we describe the ongoing integration of LUH3 spatial land-use data which includes Mapbiomas, and how the recent update of the FAO calibration data helps improve historical patterns. Finally, we briefly describe the roadmap of planned model development leading to the release of the first version of MAgPIE-Brazil. The challenges and solutions found are useful for further improvement of the model and serve as an example for development or other national model versions.