Vegetation Masking of Remote Sensing Data Aids Machine Learning for Soil Fertility Prediction

Author:

Winzeler Hans Edwin1ORCID,Mancini Marcelo23ORCID,Blackstock Joshua M.456ORCID,Libohova Zamir4ORCID,Owens Phillip R.4ORCID,Ashworth Amanda J.7ORCID,Miller David M.2,Silva Sérgio H. G.3ORCID

Affiliation:

1. Department of Mathematics, University of Texas, 411 S Nedderman Dr., Arlington, TX 76019, USA

2. Department of Crop, Soil, and Environmental Sciences, University of Arkansas, 465 Agriculture Building, Fayetteville, AR 72701, USA

3. Department of Soil Science, Universidade Federal de Lavras, Campus Universitário, Caixa Postal 3037, Lavras 37200-900, MG, Brazil

4. Dale Bumpers Small Farms Research Center, Agricultural Research Service, United States Department of Agriculture, 6883 AR-23, Booneville, AR 72927, USA

5. Center for Advanced Spatial Technologies, University of Arkansas, 227 N. Harmon Av., Fayetteville, AR 72701, USA

6. Department of Geosciences, University of Arkansas, 340 N. Campus Dr., Fayetteville, AR 72701, USA

7. Poultry Production and Product Safety Research Unit, University of Arkansas, Agricultural Research Service, United States Department of Agriculture, O-303 Poultry Science Center, Fayetteville, AR 72701, USA

Abstract

Soil nutrient content varies spatially across agricultural fields in hard-to-predict ways, particularly in floodplains with complex fluvial depositional history. Satellite reflectance data from the Sentinel-2 (S2) mission provides spatially continuous land reflectance data that can aid model development when used with point observations of nutrients. Reflectance from vegetation is assumed to obstruct land reflectance of bare soil, such that researchers have masked vegetation in models. We developed a routine for masking vegetation within Google Earth Engine (GEE) using Random Forest classification for iterative application to libraries of S2-images. Using gradient boosting, we then developed soil nutrient models for surface soils at a 250-ha agricultural site using S2 images. Soils were sampled at 2145 point locations to a 23-cm depth and analyzed for Ca, K, Mg, P, pH, S, and Zn. Results showed that masking vegetation improved model performance for models from subsets of the data (80% of samples used for model development, 20% validation), but full data sets did not require masking to achieve accuracy. Models of Ca, K, Mg, and S were successful (validation R2 > 0.60 to 0.96), but models for pH, P, and Zn failed. Bare soil composite images from S2 data are helpful in predicting soil fertility in low-relief floodplains.

Funder

National Institute of Food and Agriculture Award

SCINet project of the USDA Agricultural Research Service

Publisher

MDPI AG

同舟云学术

1.学者识别学者识别

2.学术分析学术分析

3.人才评估人才评估

"同舟云学术"是以全球学者为主线,采集、加工和组织学术论文而形成的新型学术文献查询和分析系统,可以对全球学者进行文献检索和人才价值评估。用户可以通过关注某些学科领域的顶尖人物而持续追踪该领域的学科进展和研究前沿。经过近期的数据扩容,当前同舟云学术共收录了国内外主流学术期刊6万余种,收集的期刊论文及会议论文总量共计约1.5亿篇,并以每天添加12000余篇中外论文的速度递增。我们也可以为用户提供个性化、定制化的学者数据。欢迎来电咨询!咨询电话:010-8811{复制后删除}0370

www.globalauthorid.com

TOP

Copyright © 2019-2024 北京同舟云网络信息技术有限公司
京公网安备11010802033243号  京ICP备18003416号-3