Assessing the Variability of Corn and Soybean Yields in Central Iowa Using High Spatiotemporal Resolution Multi-Satellite Imagery

Abstract

The utility of remote sensing data in crop yield modeling has typically been evaluated at the regional or state level using coarse resolution (>250 m) data. The use of medium resolution data (10–100 m) for yield estimation at field scales has been limited due to the low temporal sampling frequency characteristics of these sensors. Temporal sampling at a medium resolution can be significantly improved, however, when multiple remote sensing data sources are used in combination. Furthermore, data fusion approaches have been developed to blend data from different spatial and temporal resolutions. This paper investigates the impacts of improved temporal sampling afforded by multi-source datasets on our ability to explain spatial and temporal variability in crop yields in central Iowa (part of the U.S. Corn Belt). Several metrics derived from vegetation index (VI) time-series were evaluated using Landsat-MODIS fused data from 2001 to 2015 and Landsat-Sentinel2-MODIS fused data from 2016 and 2017. The fused data explained the yield variability better, with a higher coefficient of determination (R2) and a smaller relative mean absolute error than using a single data source alone. In this study area, the best period for the yield prediction for corn and soybean was during the middle of the growing season from day 192 to 236 (early July to late August, 1–3 months before harvest). These findings emphasize the importance of high temporal and spatial resolution remote sensing data in agricultural applications.

Using a Crop Modeling Framework for Precision Cost-Benefit Analysis of Variable Seeding and Nitrogen Application Rates

“…Additionally, studies have also attempted to quantify optimum site-specific seed densities (Licht et al, 2017), which may represent a more economically impactful management change in many cropping systems compared to changes in nutrient applications. Variable rate zones defining different application rates have been generated using precision agriculture data sources including yield monitor maps (Adamchuk et al, 2004;Basso et al, 2016;Maestrini and Basso, 2018), remotely sensed data (Hong et al, 2006;Basso et al, 2016;Gao et al, 2018;Jin et al, 2019), gridded soil sampling (Fleming et al, 2000), digital soil maps (Bobryk et al, 2016), topography (Long et al, 2015;Walters et al, 2017), and real-time optical sensors (Raun et al, 2002;Tremblay et al, 2009;Kitchen et al, 2010;Stefanini et al, 2018).…”

A Precision Agriculture Approach for Durum Wheat Yield Assessment Using Remote Sensing Data and Yield Mapping

“…During the same period many studies have been published using satellite imagery to estimate crop parameters and yields [17][18][19], many of these using empirical relationships between yields and various vegetation indices (VIs) with limited applicability to different areas or years [9], especially in the recent era of prolific satellite data availability [20].…”

“…In the last years many studies have focused on the use of medium resolution satellite data (10-100 m) for yield estimation at broader spatial resolution (local, regional, country scales) even for long-term yield series analysis [20][21][22][23][24][25][26]. While studies are conducted by the use of very high resolution imagery [27][28][29][30] to identify within-field variability of crop growth and yield and for the definition of management zones, few [31] have used Sentinel-2 to provide an insight into field productivity variation for better future management [32][33][34].…”

Crop Yield Estimation Using Time-Series MODIS Data and the Effects of Cropland Masks in Ontario, Canada

“…Later on, data from the moderate resolution imaging spectroradiometer (MODIS) has been used more often because of better spatial, spectral, and radiometric resolutions [1,21]. A recent research trend is to merge these coarse resolution satellite data with high-spatial low-temporal resolution satellite data (e.g., Landsat data) to generate high resolution time-series data for field level studies [17,[27][28][29]. Data volume and computation-resource requirement are high if this method is applied at the regional scale and over a long time period, although mature cloud computing techniques might provide a solution.…”