Enhancing Crop Yield Prediction Utilizing Machine Learning on Satellite-Based Vegetation Health Indices

Abstract

Accurate crop yield forecasting is essential in the food industry’s decision-making process, where vegetation condition index (VCI) and thermal condition index (TCI) coupled with machine learning (ML) algorithms play crucial roles. The drawback, however, is that a one-fits-all prediction model is often employed over an entire region without considering subregional VCI and TCI’s spatial variability resulting from environmental and climatic factors. Furthermore, when using nonlinear ML, redundant VCI/TCI data present additional challenges that adversely affect the models’ output. This study proposes a framework that (i) employs higher-order spatial independent component analysis (sICA), and (ii), exploits a combination of the principal component analysis (PCA) and ML (i.e., PCA-ML combination) to deal with the two challenges in order to enhance crop yield prediction accuracy. The proposed framework consolidates common VCI/TCI spatial variability into their respective subregions, using Vietnam as an example. Compared to the one-fits-all approach, subregional rice yield forecasting models over Vietnam improved by an average level of 20% up to 60%. PCA-ML combination outperformed ML-only by an average of 18.5% up to 45%. The framework generates rice yield predictions 1 to 2 months ahead of the harvest with an average of 5% error, displaying its reliability.

Augmenting agroecosystem models with remote sensing data and machine learning increases overall estimates of nitrate-nitrogen leaching

“…There are several possible answers to this problem, including insufficient sample size or consecutive years of data; the natural variation in yearly weather patterns; or a lack of sites with similar enough features. While identifying the best yield prediction model was not the explicit goal of this study, this is a research area we would like to explore more in the future with more field sites, years of data, and high-resolution images (Pham et al 2022). This also illustrates the need for more high quality, long-term, calibrated yield monitor data to identify within-field productivity trends (Bunselmeyer and Lauer 2015) and improve ecosystem service assessments.…”

Precipitation and Soil Moisture Spatio-Temporal Variability and Extremes over Vietnam (1981–2019): Understanding Their Links to Rice Yield

“…The results show correlation coefficients at different magnitudes with the maximum and minimum values being ∼0.85 and ∼0.34 at a 95% confidence level, respectively, and lower values were found in the Northwest (1), Central Coast (4, 5), and Central Highlands (6) areas. Areas exhibiting lower correlation coefficients occupied either the mountainous areas (1,6) or the coastal areas (4, 5), see Figure 1. This is probably because high relief characteristics in mountainous areas reduce the ability of soil to store water or water sources other than precipitation, e.g., seawater, is contributed in coastal areas, both of which result in lower correlation between precipitation and soil moisture.…”

“…Vietnam is located in southeast Asia and is well-known for its rice production, where it lies amongst the top three rice producers globally [1]. Domestically, its rice sustains a livelihood of more than 96 million people.…”

“…This is because rice yield is vulnerable to drought/flood stress due to its shallower rooting than other crops [6]. Therefore, the knowledge of precipitation and soil moisture variability is essential to inform policy and adaptation strategies in order to prepare for climate extremes to minimize impacts on Vietnamese rice productivity [1].…”

Improving the quality of coffee yield forecasting in Dak Lak Province, Vietnam, through the utilization of remote sensing data

“…With the Machine Learning method, climate data (Dinh et al 2022), soil nutrient composition, and the plant indexes in the growing area (Kouadio et al 2018) were often used as predictors of coffee yield. Various Machine Learning algorithms, such as artificial neural networks and regressions, have been utilized to develop yield forecasting models (Pham et al 2022). To effectively predict coffee yield, it is essential to consider the complex interplay between various environmental factors and geographical variables.…”