Multistage Soybean Biomass Inversion Models and Spatiotemporal Analyses considering Microtopography at the Sub-Field Scale. Issue 1 (2nd January 2019)

• Record Type: Journal Article
• Authors: Xu, Mengyuan; Zhang, Xinle; Meng, Linghua; Liu, Huanjun; Pan, Yue; Qiu, Zhengchao; Yang, Haoxuan; Zhang, Zhanqian
• Is Part Of: Canadian journal of remote sensing. Volume 45:Issue 1(2019)
• Page Start: 1
• Page End: 15
• Publication Date: 2019-01-02
• Subjects: Remote sensing -- Periodicals; 621.367805
• DOI: 10.1080/07038992.2019.1594176
• Languages: English
• ISSNs: 0703-8992

Abstract

Abstract: Crop biomass is an agricultural indicator of productivity, and knowledge of the temporal and spatial variation in biomass in different topography is critical for the application of precision management techniques. This study integrates microtopography with multitemporal remote sensing observations to reveal biomass- and yield-limiting variables. Three SPOT-6 high spatial resolution images and six topographic variables were combined to model the spatial variation in soybean biomass at multiple stages during the growing season. The results showed the following. ( i ) The multiple regression model for biomass estimation that combines topographic variables with vegetation indices can achieve higher accuracy than a vegetation index model. ( ii ) Biomass varied dramatically with topography during the growing season. ( iii ) Microtopographic variables, such as curvature and slope, had distinctive impacts on crop conditions over a growing season. Early in the growing season, sunny upper slopes produced more biomass than shady lower slopes, whereas this trend reversed over the season. Gentle concave slopes (–1.2 m −1 < curvature < 0 m −1 ) showed greater productivity later in the season, while slopes with high concavity (curvature < –1.2 m −1 ) or high convexity (curvature > 0.75 m −1 ) suppressed crop growth. These conclusions can be used directly to precisely manage nutrients and water applications.