Identifying soil managements via hyperspectral remote sensing: A method for determining allocation of manure-based nutrients within potential manuresheds

Authors: Flynn, Kyle; JEONG, JAEHAK - Texas Agrilife Research; KAN, EUNSUNG - Texas Agrilife Research; MUIR, JAMES - Texas Agrilife Research; Polley, Herbert

Submitted to: ASA-CSSA-SSSA Annual Meeting Abstracts
Publication Type: Abstract Only
Publication Acceptance Date: 9/1/2021
Publication Date: N/A
Citation: N/A

Interpretive Summary:

Technical Abstract: Manuresheds – the croplands surrounding livestock production operations that could benefit from the redistribution of livestock manure nutrients – require spatial knowledge of management decisions by cropland managers in close proximity to confined animal operations that could benefit from the allocation of manure-based nutrients. However, identifying these nearby fields can be time and labor intensive as it often requires interaction with the cropland managers to identify potential manure placement. The advancements among remote sensing platforms provides opportunity to determine soil management decisions remotely. Utilizing reflectance data provides opportunity to determine cropland soils that could uptake manure in place of other soil managements (i.e. synthetic fertilizers). Therefore, accurate quantification of various soil amendments (e.g. manure, biochar, or synthetic fertilizer) could aid in soil management mapping and improved manuresheds. Thus, in a controlled greenhouse setting with soils (n=234) from two locations (Temple & Stephenville, TX) we amended the soils with either dairy manure, biochar, or synthetic fertilizer while cultivating bermuda grass (Cynodon dactylon). We employed the use of a visible-near infrared sensor measuring reflectance from 400-2500nm at 1nm wavelength intervals on two dates at the beginning and end of the experiment (n=468). The categorical variable of amendment type served as the independent variable and reflectance data served as the dependent variables, the collected data were subject to partial least squares discriminant analysis (PLS-DA). We then convolved the spectral reflectance of the visible-near infrared sensor to approximate the spectral bands (220 band; 6.5-10nm bandwidths) of the soon-to-launch (2021) Environmental Monitoring and Analysis Program (EnMAP) satellite-based platform. The newly convolved bands were then subject to the same PLS-DA to determine potential for the application of a satellite-platform to determine soil managements. The findings suggest that both in-situ and satellite-based remote sensing can determine soil management decisions, thereby potentially improving the aims of the manureshed concept.