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

Vietnam, one of the three leading rice producers globally, has recently seen an increased threat to its rice production emanating from climate extremes (floods and droughts). Understanding spatio-temporal variability in precipitation and soil moisture is essential for policy formulations to adapt and cope with the impacts of climate extremes on rice production in Vietnam. Adopting a higher-order statistical method of independent component analysis (ICA), this study explores the spatio-temporal variability in the Climate Hazards Group InfraRed Precipitation Station's (CHIRPS) precipitation and the Global Land Data Assimilation System's (GLDAS) soil moisture products. The results indicate an agreement between monthly CHIRPS precipitation and monthly GLDAS soil moisture with the wetter period over the southern and South Central Coast areas that is latter than that over the northern and North Central Coast areas. However, the spatial patterns of annual mean precipitation and soil moisture disagree, likely due to factors other than precipitation affecting the amount of moisture in the soil layers, e.g., temperature, irrigation, and drainage systems, which are inconsistent between areas. The CHIRPS Standardized Precipitation Index (SPI) is useful in capturing climate extremes, and the GLDAS Standardized Soil Moisture Index (SSI) is useful in identifying the influences of climate extremes on rice production in Vietnam. During the 2016-2018 period, there existed a reduction in the residual rice yield that was consistent with a decrease in soil moisture during the same time period.

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

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.