Assessing variability of soil water balance components measured at a new lysimeter facility dedicated to the study of soil ecosystem services.

• A new lysimeter facility dedicated to studying soil ecosystem services is introduced. • Replicate weight measurements were used to assess variability of the water balance. • Lysimeters measured precipitation to a high precision (±2.4% of the mean over 120-days). • Data processing procedures added a minimal amount of uncertainty to water balance sums. • Spatial variations were not driven by intrinsic soil core differences. Diversifying annual crop rotations is postulated to promote soil ecosystem services through reduction of greenhouse gas emissions, control of soil evaporation, moderation of soil freeze-thaw cycles, increases in soil organic carbon, and reduced nutrient leaching, all of which are processes that are impacted by the water balance. A newly-installed facility with large monolithic weighing soil lysimeters dedicated to studying changes in soil ecosystem services induced by rotation diversification was installed in June 2016 in Elora, ON, Canada. The high-precision lysimeters, nine for each of two different soils (silt loam, loamy sand) are a closed system to fully evaluate water balance components. In the first year of operation, the lysimeters were not subject to any diversification treatments. This provided an opportunity to study water budget components (precipitation, P ; evapotranspiration, ET ; change is soil storage, Δ S ; drainage, D) measured by the 9 replicate lysimeters for each soil type. The objectives of this study were to 1) determine the total uncertainty of P , ET , Δ S and D induced by data smoothing and gap filling data procedures; and 3) characterize the spatial variability of water budget components for the baseline year. Cumulative sums of water balance variables from data processing procedures found that the adaptive window and adaptive threshold (AWAT) data smoothing method introduced minimal errors (0.2 to 1.9%). Filling missing data gaps using data from replicate lysimeters, i.e., lysimeters of the same soil type and agronomy, introduced errors of only 0.3 to 0.5% for P and ET (120-day sums). There was greater uncertainty for D and Δ S (10.8% and 2.1%, respectively) because of the variability in D. P showed a small degree of spatial variation across all 18 lysimeters, with variations being 1.5% of the mean cumulative sums when using data with no gaps, which demonstrated the high precision of the lysimeter measurements. Spatial variability of evapotranspiration (ET) was higher (6.1% for the silt loam and 5.7% for the loamy sand), but within the range of variability estimated for similar lysimeter facilities. The differences in plant cover on individual lysimeters were the main source of variation for ET and Δ S. Results inform future comparisons of water budget components for other lysimeter sites and demonstrate the suitability of this facility for examining the complex impacts of crop diversification on soil ecosystem services, providing a basis for future evaluations of water balance components. [ABSTRACT FROM AUTHOR]