Predicting spatial and temporal variability in crop yields: an inter-comparison of machine learning, regression and process-based models

Abstract

Pervious assessments of crop yield response to climate change are mainly aided with either process-based models or statistical models, with a focus on predicting the changes in average yields, whilst there is growing interest in yield variability and extremes. In this study, we simulate US maize yield using process-based models, traditional regression model and a machine-learning algorithm, and importantly, identify the weakness and strength of each method in simulating the average, variability and extremes of maize yield across the country. We show that both regression and machine learning models can well reproduce the observed pattern of yield averages, while large bias is found for process-based crop models even fed with harmonized parameters. As for the probability distribution of yields, machine learning shows the best skill, followed by regression model and process-based models. For the country as a whole, machine learning can explain 93% of observed yield variability, followed by regression model (51%) and process-based models (42%). Based on the improved capability of the machine learning algorithm, we estimate that US maize yield is projected to decrease by 13.5% under the 2 • C global warming scenario (by~2050 s). Yields less than or equal to the 10th percentile in the yield distribution for the baseline period are predicted to occur in 19% and 25% of years in 1.5 • C (by~2040 s) and 2 • C global warming scenarios, with potentially significant implications for food supply, prices and trade. The machine learning and regression methods are computationally much more efficient than process-based models, making it feasible to do probabilistic risk analysis of climate impacts on crop production for a wide range of future scenarios.

Climate change as a driver of food insecurity in the 2007 Lesotho-South Africa drought

“…Maize production in Lesotho and South Africa is mainly rainfed 7 , which results in precipitation during the JFM maize growing season being correlated with maize production (ϱ = 0.69 in South Africa and ϱ = 0.56 in Lesotho). A relationship between precipitation and maize yields is also found in other regions globally, such as Sub-Saharan Africa 46 , the USA 47 and China 48 .…”

Modelling global impacts of climate variability and trend on maize yield during 1980–2010

“…Both statistical and process‐based crop models are the major tools used for assessing climate impacts on crop yield (Osborne and Wheeler, 2013; Asseng et al ., 2015; Dawson et al ., 2016; Deryng et al ., 2016; Iizumi and Ramankutty, 2016; Anderson et al ., 2018; Feng and Hao, 2020; Leng and Hall, 2020), but each of them has its own strength and weakness (Lobell and Asseng, 2017; Rötter et al ., 2018). To date, limited studies have been conducted to compare climate impacts on crop yield between the two methods (White et al ., 2011; Challinor et al ., 2014; Lobell and Asseng, 2017).…”

How Can We Realize Sustainable Development Goals in Rocky Desertified Regions by Enhancing Crop Yield with Reduction of Environmental Risks?

“…Second, the large differences occurred between existing model outputs for simulating crop yield, and the relative error in some cases surpassed 50% [35]. Third, the reported accuracy for traditional regression and process-based models could only explain 51% and 42% of observed yield variability, respectively [77]. The technology of machine learning, including ANN, can counter these deficiencies to some extent.…”