The global food system's fragility is highlighted by crises like the COVID-19 pandemic and the war in Ukraine, exacerbating existing climate change impacts on food production. Shortages caused by extreme weather or conflict often have regional origins but global consequences, impacting food prices and consumption worldwide, particularly in vulnerable regions. Improving global food system resilience is crucial. Satellite-based Earth observations offer a synoptic view of environmental change, enabling timely monitoring of crop growing conditions. Recent advancements in satellite remote sensing have significantly improved the accuracy and timeliness of national-scale cereal yield estimations, providing forecasts one to two months before harvest. This enhanced forecasting capability creates opportunities to enhance the stability of global food production, supply, and prices. The complementary crop calendars of the Northern and Southern Hemispheres suggest that reliable, timely information on crop harvests in one hemisphere could influence production in the other, mitigating price volatility. This study quantifies the market stabilization benefits of this cross-hemispheric response mechanism using satellite-based forecasting data.

The literature highlights the decades-long use of remote sensing data for agricultural monitoring, but research connecting this data to improved decision-making and societal outcomes has lagged. Studies have demonstrated the potential of remote sensing for yield forecasting, using methods such as biophysical crop-simulation models and statistical regression approaches. While these methods show promise for individual countries, research examining cross-hemispheric interactions and its impact on global food security has been limited. This study aims to bridge that gap by systematically examining the "tele-connected" cross-hemispheric interactions.

This study quantifies the market stabilization benefits of the cross-hemispheric response mechanism using three case studies: (1) a better-than-normal wheat harvest in Russia and Ukraine in 2008; (2) a very poor wheat harvest in these regions in 2012; and (3) a very poor soybean harvest in southern Brazil in 2012. Remote sensing-based forecasting information for these cases, accurate 1-2 months before harvests, was integrated with a computable general equilibrium (CGE) modeling approach using the GTAP database with a land allocation module. The effects of real yield shocks were first assessed using historical data from the FAO and Companhia Nacional de Abastecimento. Then, supply response scenarios were established using the remote-sensing forecasts and compared with the real shocks to quantify the impact of forecasting on market stabilization. The remote sensing data used included MODIS Surface Reflectance for wheat yield forecasting (using a generalized regression-based model) and a machine-learning-based classification model for soybean cover mapping in South America. The CGE models used involved 141 customs territories/regions and 65 industrial sectors, aggregated into smaller numbers for the analyses. The models incorporated land allocation, production functions with various elasticities (obtained from the GTAP database and relevant literature), and consumer demand functions. Robustness checks were performed against different elasticity parameters.

Remote sensing accurately forecasted wheat production changes in Russia and Ukraine (74-78% accuracy in 2008, 56-73% in 2012) and soybean production in southern Brazil (44% loss vs. actual 35%). Abnormal wheat harvests resulted in significant price changes in importing regions (e.g., Japan: -29% in 2008, +24% in 2012). Using satellite forecasts, price volatility was reduced (e.g., Japan: -21% in 2008, +15% in 2012). The poor soybean harvest in Brazil caused price increases in major importing regions, but the forecasts helped mitigate this, in some cases even causing price decreases (e.g., Mexico: -6.4%). Forecasting triggered cross-hemispheric responses: in 2008, better-than-normal wheat harvests led to reduced wheat production in the Southern Hemisphere (2.3-2.2% of global wheat export), while in 2012, poor harvests prompted increased wheat production in the South (1.8-2.3% of world export). The soybean forecast in Brazil led to increased soybean production in the US and Canada (4.8% of global soybean export). Overall, the satellite-based forecasts mitigated price fluctuations across multiple countries. Robustness checks revealed that the results were qualitatively robust, even with variations in elasticity parameters.

The results demonstrate that timely and accurate satellite-based crop forecasting can significantly stabilize agricultural markets by triggering cross-hemispheric production responses. This mechanism offers three main benefits: (1) market stabilization, reducing price volatility and risk premiums; (2) improved resource allocation efficiency; and (3) decreased opportunities for speculative activities. This approach leverages the inherent seasonal lag between the hemispheres to mitigate supply shocks effectively.

This research shows that satellite-based crop forecasting can be a powerful tool to improve global food security. Timely, accurate forecasts trigger cross-hemispheric production responses that stabilize markets and mitigate price shocks. Future research should focus on improving forecasting accuracy (to above 90% two months before harvest), examining the role of speculation, and using recursive semi-annual models to capture intra-annual price volatility more precisely. Policy implications include minimizing trade restrictions to facilitate the cross-hemispheric response mechanism and investing in research and development to enhance forecasting accuracy and accessibility.

The study's limitations include assumptions of flexible crop choices among farmers in the opposite hemisphere, ignoring potential constraints of crop rotation and other conditions, underestimation of speculative activities impact and the number of responsive farmers. The annual timeframe of the CGE model simplifies the analysis; a recursive semi-annual model would offer more nuanced insights into intra-annual price volatility.