Location: Soil and Water Management ResearchTitle: Soil water sensing for climate change studies; Applicability of COSMOS and local sensor networks
|Evett, Steven - Steve|
|Brauer, David - Dave|
|FRANZ, TRENTON - University Of Nebraska|
Submitted to: Meeting Abstract
Publication Type: Abstract Only
Publication Acceptance Date: 12/9/2014
Publication Date: 12/9/2014
Citation: Evett, S.R., Schwartz, R.C., Brauer, D.K., Franz, T., Ruthardt, B.B., Copeland, K.S. 2014. Soil water sensing for climate change studies; Applicability of COSMOS and local sensor networks [abstract].
Technical Abstract: Soil water sensors are used to characterize water content in the near-surface, the root zone and below for agricultural and ecosystem management, but only a few are capable of sensing soil volumes larger than a few hundred liters. Scientists with the USDA-ARS Conservation & Production Research Laboratory, Bushland, Texas, evaluated and compared three soil water sensing systems with each other and with precipitation and irrigation amounts measured with a large weighing lysimeter. The three sensor systems were: 1) the Cosmic Ray Soil Moisture Observing System (COSMOS), which responds to surface soil water content changes in a circular area of radius up to several hundred meters; 2) the neutron probe (NP), used in a network of eight access tubes spaced around the lysimeter to take readings centered at depths from 0.10 to 2.30 m in depth increments of 0.20 m, and 3) electromagnetic (EM) soil water sensors (model CS655, Campbell Scientific, Inc., Logan, Utah) that each sense only a few hundred cubic centimeters. The CS655 sensors were used in a wireless sensor network to interrogate larger volumes of soil over the 0-0.30 m depth range at the eight NP sites. A large precision weighing lysimeter was used to measure precipitation, irrigation and soil water storage amounts. The weighing lysimeter had a calibrated accuracy of 0.04 mm (<0.01 inch). Uncorrected COSMOS data were well correlated (r2=0.87) with 0-0.30 m (1 ft) water content and storage as measured by the field-calibrated CS655 sensors. COSMOS was more sensitive to increases in soil water from rainfall compared with soil water increases due to subsurface drip irrigation at 0.30-m depth. COSMOS water content data were biased upward by green, living vegetation (corn in 2013 and sorghum in 2014) and by atmospheric humidity increases, including those due to rapid evaporation from the soil surface after a wetting event. When used to calculate soil water storage, the effective depth of the COSMOS measurement, which is calculated from a relationship based on neutron transport simulations, resulted in storage values that were poorly correlated with measured soil water storage in this study. However, assuming that the effective depth was constant at 0.30 m depth resulted in better correlation with CS655 measured soil water storage. Correcting COSMOS data for clay lattice water, water vapor (humidity of the air), soil organic matter and dry and wet biomass did not result in better correlations with other measurements. Indeed, the coefficient of determination for corrected COSMOS data versus 0-0.30 m CS655 data decreased to 0.84 from the previous 0.87 for uncorrected COSMOS data. Depth sensitivity studies showed that COSMOS was progressively better correlated with data that represented increasing depth ranges from 0-0.075 to 0-0.175 m to 0-0.275 m. The wireless CS655 sensor system worked very well, providing timely information that correlated well with weighing lysimeter soil water storage data. The standard deviation of soil water storage for the eight CS655 sites was similar to that for the eight collocated NP sites. The CS655 network was the only system compared that provided bulk electrical conductivity data that would be useful for salinity management. Data from the COSMOS system were not accurate enough for irrigation scheduling, and are not responsive to soil salinity. The latter is largely a positive attribute since it means that the COSMOS data do not have to be corrected for soil bulk electrical conductivity. Data from the neutron probe were accurate but labor intensive, and the neutron probe lacked both automation and wireless data transfer.