Monitoring of sugar beet growth indicators using wide-dynamic-range vegetation index (WDRVI) derived from UAV multispectral images

Highlights

• WDRVI improved the accuracy of Beet LAI, FWL, and FWR predictions with Vis.

• WDRVI1 (α = 0.01) provides the most accurate estimation of beet growth indicators.

• Drone-mount sensor have great potential for estimating plot-level beet growth indicators.

Abstract

The normalized vegetation index (NDVI) is widely used to monitor the spatial, temporal, physiological, and biophysical characteristics of vegetation. However, if the ground biomass is high, the NDVI becomes rapidly saturated. In the leaf area index (LAI) range of 2 to 6, reflectance in the near-infrared band is significantly higher than that in the red band. When reflectance in the near-infrared band exceeds 40%, the contribution of that reflectance to the NDVI is low. Here, we applied a weight coefficient, α (range 0.05 to 0.5), to the near-infrared band, thus developing a wide-dynamic-range vegetation index (WDRVI). We calculated (α * NIR-RED)/(α * NIR + RED) values; these emphasized the Fresh Weight of Beet LAI, the Fresh Weight of Leaves (FWL), and the Fresh Weight of Roots (FWR). We sought a vegetation index that optimally monitored early-stage sugar beet growth. The WDRVI sensitivities using various α coefficients were higher than those of the NDVI in the LAI range of 2 to 6. The determination coefficients (r² values) of the sugar beet LAI, FWL, and FWT models established using the WDRVI1 were 0.957, 0.950, and 0.963, respectively. Using the WDRVI1 index model to estimate the accuracy of beet growth indicators can improve 1.05% to 5.07%. Use of the WDRVI reduces beet growth indicator saturation when the ground biomass is high, enhancing the accuracy of growth monitoring. This is useful as the sugar beet has huge ground biomass.

Introduction

Sugar beet is an important cash crop in northern China and the second-largest sugar crop in the country. Beet is grown principally in Inner Mongolia, Xinjiang, and Heilongjiang. The beet biomass can reach more than 60 t/ha, thus much greater than those of corn, wheat, and other crops. However, few reports on the use of multispectral platforms to monitor sugar beet growth have appeared (Li et al., 2017a, Li et al., 2017b). With the expansion of beet planting in Inner Mongolia, the establishment of a precise sugar beet management system has become necessary. Efficient non-destructive monitoring of crop growth is necessary for accurate crop management. The Leaf Area Index (LAI), Fresh Weight of Leaves (FWL), and Fresh Weight of Roots (FWR) are important parameters used to monitor beet canopy structure and growth.

In the time since the initial development of remote sensing technology in the 1950s, satellite and manned aircraft platforms, and ground spectral equipment have been used to monitor crop growth. However, certain issues remain in play. Satellite platforms are constrained by both altitude and orbit, and they do not afford the spatial, temporal, or spectral resolution required for growth monitoring (White et al., 2009, Ortiz et al., 2011). Satellite views are constrained by clouds and suspended particles. The operating costs of manned platforms are high; personnel require advanced flight training. Ground spectral equipment is cumbersome; the operating speed is only about 0.56 m/s, and the crop canopy is inevitably damaged. Newer unmanned aerial vehicle (UAV) technology affords cost-effective crop growth monitoring at high spatial, temporal, and spectral resolution. Enciso J A found that coefficient of determination of 0.72 was observed between canopy cover estimated with the UAV and leaf area index measured with the ceptometer (Enciso et al., 2019). Drone sensors are either multispectral or hyperspectral cameras facilitating LAI monitoring, biomass estimation, and yield prediction. Earlier fixed-wing UAVs were fast and covered large areas. ZHANG found that UAV images with high spatiotemporal resolution can be combined with Aqua Crop model to estimate wheat growth (Zhang et al., 2019). The UAV-based technique was able to detect and count citrus trees with high precision (99.9%) in an orchard of 4931 trees and estimate tree canopy size with a high correlation (R = 0.84) with the manual collected data (Arnpatzidis et al., 2019). However, it was difficult to obtain high-quality spectral information, and take-off usually required manual assistance or the use of a stand. Additionally, landings were generally hard, compromising safety. Few crop-monitoring studies employed fixed-wing UAV platforms. Unlike fixed-wing aircraft, multi-rotor UAVs are stable and controllable, fly at low altitudes, are simple to operate, and pose no problems in terms of take-off or landing. Drones now find many applications in agricultural crop growth monitoring. RGB images captured by UAV, and used to estimate the canopy projected area of individual trees, proved to be the best of the options. This was shown by the high correlation (R = 0.85) between this area and the fruit load (Uribeetxebarria et al., 2019). Devia monitored the agronomic parameters of rice using a seven-point index (Devia et al., 2019). Duan collected a great deal of information on rice growth using a UAV vegetation index and spectral characteristics, and successfully predicted the extent of rice heading (Duan et al., 2019). Multispectral sensors on UAV platforms have been successfully used to monitor the yields and above-ground biomasses of grass, wheat, and sunflowers (Elmore et al., 2000, Honda et al., 2014). Liu found that the use of RVI obtained by drone can monitor the aboveground biomass of winter oilseed rape (Liu et al., 2019). Over the past 30 years, vegetation indices have been widely used to evaluate crop canopy structures and growth. The most widely used index is the normalized vegetation index (NDVI); this assesses the biophysical characteristics of canopies. UAV based hyperspectral images linked to a radiative transfer model can provide a promising approach for high throughput monitoring of plant nitrogen (N) status. Results suggest that Multi-LUTs of leaf area index, leaf N density and two spectral indices (MSR and MCARI/MTVI2) in winter wheat demonstrate good performance of canopy N density (CND) estimation (Li et al., 2019). Zhou found that there is a strong relationship between green normalized difference vegetation index (GNDVI), NDVI and yield data for multiple potato varieties (r = 0.77–0.90) (Zhou et al., 2016). The NDVI-derived LAI and leaf clump biomass (Rouse et al., 1974) are widely used to monitor growth, but medium and high biomasses saturate the NDVI (Sellers, 1985, Gitelson et al., 2002). Thus, a wide-range dynamic vegetation index (WDRVI) was developed; the WDRVI performs weighted calculations using NVDI near-infrared band reflectance, reducing the weights of such bands under medium-to-high biomass conditions. This enhances linearity and reduces saturation, facilitating accurate growth monitoring (Huete et al., 2002, Asrar et al., 1984). Here, we used an improved WDRVI to monitor the early-stage growth of sugar beet. We adjusted the weight coefficient of the near-infrared band and identified an optimal vegetation index.