High temporal resolution of leaf area data improves empirical estimation of grain yield

Abstract

Empirical yield estimation from satellite data has long lacked suitable combinations of spatial and temporal resolutions. Consequently, the selection of metrics, i.e., temporal descriptors that predict grain yield, has likely been driven by practicality and data availability rather than by systematic targetting of critically sensitive periods as suggested by knowledge of crop physiology. The current trend towards hyper-temporal data raises two questions: How does temporality affect the accuracy of empirical models? Which metrics achieve optimal performance? We followed an in silico approach based on crop modelling which can generate any observation frequency, explore a range of growing conditions and reduce the cost of measuring yields in situ. We simulated wheat crops across Australia and regressed six types of metrics derived from the resulting time series of Leaf Area Index (LAI) against wheat yields. Empirical models using advanced LAI metrics achieved national relevance and, contrary to simple metrics, did not benefit from the addition of weather information. This suggests that they already integrate most climatic effects on yield. Simple metrics remained the best choice when LAI data are sparse. As we progress into a data-rich era, our results support a shift towards metrics that truly harness the temporal dimension of LAI data.

The role of asymmetric information in multi-peril picture-based crop insurance: Field experiments in India

“…This will be true especially when heterogeneity in phenology between fields and seasons is large due to differences in farm management practices, crop varietal choices, and weather conditions. We also find that insurance index performance can be improved further by combining LAI and weather predictor variables, which we attribute the ability of weather data (in particular temperature predictors) to capture crop yield losses associated with deficient grain filling or pollination that would not be fully captured by changes in LAI alone (Waldner et al, 2019).…”

Prediction of winter wheat yield at county level in China using ensemble learning

“…The key to solving this problem is to produce more accurate crop classification as a mask for high-resolution data. Monthly NDVI and climate data may ignore some crop growth and weather information, and higher temporal resolution data can more accurately reflect crop productivity (Waldner et al, 2019). Due to the influence of soil background, NDVI tends to be saturated at high canopy density, so it will not further explain the difference in biomass (Schwalbert et al, 2020).…”

Towards a massive sentinel-2 LAI time-series production using 2-D convolutional networks

“…LAI is a biophysical parameter that measures the total area of leaves per unit ground area and is directly correlated with the amount of intercepted light by the plant. This parameter has many uses such as the prediction of photosynthetic primary production, monitoring crop growth and yield estimation (Waldner et al, 2019). Moreover, the LAI is required by many global models of climate (Bonan et al, 2002), ecosystem productivity and ecology (Asner et al, 2003;Running and Coughlan, 1988;Sellers et al, 1997;Yan et al, 2016).…”