No abstract

Chlorophyll is the most important component of crop photosynthesis, and the reviving stage is an important period during the rapid growth of winter wheat. Therefore, rapid and precise monitoring of chlorophyll content in winter wheat during the reviving stage is of great significance. The satellite-UAV-ground integrated inversion method is an innovative solution. In this study, the core region of the Yellow River Delta (YRD) is used as a study area. Ground measurements data, UAV multispectral and Sentinel-2A multispectral imagery are used as data sources. First, representative plots in the Hekou District were selected as the core test area, and 140 ground sampling points were selected. Based on the measured SPAD values and UAV multispectral images, UAV-based SPAD inversion models were constructed, and the most accurate model was selected. Second, by comparing satellite and UAV imagery, a reflectance correction for satellite imagery was performed. Finally, based on the UAV-based inversion model and satellite imagery after reflectance correction, the inversion results for SPAD values in multi-scale were obtained. The results showed that green, red, red-edge and near-infrared bands were significantly correlated with SPAD values. The modeling precisions of the best inversion model are R2 = 0.926, Root Mean Squared Error (RMSE) = 0.63 and Mean Absolute Error (MAE) = 0.92, and the verification precisions are R2 = 0.934, RMSE = 0.78 and MAE = 0.87. The Sentinel-2A imagery after the reflectance correction has a pronounced inversion effect; the SPAD values in the study area were concentrated between 40 and 60, showing an increasing trend from the eastern coast to the southwest and west, with obvious spatial differences. This study synthesizes the advantages of satellite, UAV and ground methods, and the proposed satellite-UAV-ground integrated inversion method has important implications for real-time, rapid and precision SPAD values collected on multiple scales.

No abstract

The objectives of this study were to explore the spatial variability of soil salinity in coastal saline soil at macro, meso and micro scales in the Yellow River delta, China. Soil electrical conductivities (ECs) were measured at 0-15, 15-30, 30-45 and 45-60 cm soil depths at 49 sampling sites during November 9 to 11, 2013. Soil salinity was converted from soil ECs based on laboratory analyses. Our results indicated that at the macro scale, soil salinity was high with strong variability in each soil layer, and the content increased and the variability weakened with increasing soil depth. From east to west in the region, the farther away from the sea, the lower the soil salinity was. The degrees of soil salinization in three deeper soil layers are 1.14, 1.24 and 1.40 times higher than that in the surface soil. At the meso scale, the sequence of soil salinity in different topographies, soil texture and vegetation decreased, respectively, as follows: depression >flatland >hillock >batture; sandy loam >light loam >medium loam >heavy loam >clay; bare land >suaeda salsa >reed >cogongrass >cotton >paddy >winter wheat. At the micro scale, soil salinity changed with elevation in natural micro-topography and with anthropogenic activities in cultivated land. As the study area narrowed down to different scales, the spatial variability of soil salinity weakened gradually in cultivated land and salt wasteland except the bare land.

Soil salinization adversely impacts crop growth and production, especially in coastal areas which experience serious soil salinization. Therefore, rapid and accurate monitoring of the salinity and distribution of coastal saline soil is crucial. Representative areas of the Yellow River Delta (YRD)—the Hekou District (the core test area with 140 sampling points) and the Kenli District (the verification area with 69 sampling points)—were investigated. Ground measurement data, unmanned aerial vehicle (UAV) multispectral imagery and Sentinel-2A multispectral imagery were used as the data sources and a satellite-UAV-ground integrated inversion of the coastal soil salinity was performed. Correlation analyses and multiple regression methods were used to construct an accurate model. Then, a UAV-based inversion model was applied to the satellite imagery with reflectance normalization. Finally, the spatial and temporal universality of the UAV-based inversion model was verified and the soil salinity inversion results were obtained. The results showed that the green, red, red-edge and near-infrared bands were significantly correlated with soil salinity and the spectral parameters significantly improved this correlation; hence, the model is more effective upon combining spectral parameters with sensitive bands, with modeling precision and verification precision of the best model being 0.743 and 0.809, respectively. The reflectance normalization yielded good results. These findings proved that applying the UAV-based model to reflectance normalized Sentinel-2A images produces results that are consistent with the actual situation. Moreover, the inversion results effectively reflect the distributions characteristic of the soil salinity in the core test area and the study area. This study integrated the advantages of satellite, UAV and ground methods and then proposed a method for the inversion of the salinity of coastal saline soils at different scales, which is of great value for real-time, rapid and accurate soil salinity monitoring applications.