We estimate the water balance by measuring incoming (precipitation) and outgoing water fluxes (stream flow). We might also estimate water loss from evapo-transpiration, which will combine weather station climate (AWS) data with chamber-based carbon flux data (LiCor), which also provide water fluxes from soil and during light measurements over vegetation. Moreover, we have water table depth loggers placed across the landscape to determine accuracy of model predictions and to relate water levels to runoff and evaporation rates.
Weather station for measuring incoming rainfall
Outgoing stream flow with flow weir to measure flow rates
Plot-scale water table depth monitoring
Input versus outflow per catchment (as %loss of incoming rainfall measured at the automated weather station (AWS))
Example for 2012: Basically, at the start of the monitoring period (very dry) most of the water stayed within the system (raising the water table), whereas after saturation more than 80% are lost as runoff
Flow rates (Q) measured at catchment outflow weirs; Example for 2012: All three sites with control (C) and treatment (T) are shown. The three sites show remarkably similar flow rate peaks and are similar between the control and treatment areas
Water table depth (WTD) in 2012 (means of four plots)
Example for 2012: Controls given with STDEV; DN denotes "do nothing" treatment. Note: Whitendale contains one very low WTD control replicate
Above: Actual evapo-transpiration rates vs. PAR at Nidderdale July (left) versus October (right) 2012. Lines are linear regressions - however, asymptotic curves will be fitted for modelling purposes. (Note: this equates to up to an estimated (back on the envelope that is) 25,000 m3 per year per catchment or less than 10% of rainfall input)
Above: Mean monthly water table depths (WTD) at the main monitoring plots across all sites for the main treatments: uncut, burnt, mown with leaving brash (LB) or brash removal (BR) for combined Sphagnum pellet additions. The red arrow separates pre- and post-management periods (i.e. mowing and burning in spring 2013). All plots lost their vegetation in spring 2013 apart from the do nothing (uncut) plots. Plots have re-grown since and additional burning and mowing happened elsewhere in the catchments. From left to right (Nidderdale - Mossdale - Whitendale).
Above: Water table depth shown for the uncut plots (left) as monthly averages over time and as overall averages (right) for the pre- versus post-management periods.
Above: Water balance shown as the loss as flow volume measured at the flow weirs as a percentage of the total incoming rainfall, either as monthly totals (left) or the annual averages (right) for the pre- and post-management period including the subsequent management events (all burning and mowing events are highlighted in red and green, respectively). Any loss (%) difference between mown and burnt catchments greater than 10% for values of burnt flow loss and the difference between burnt and mown of greater than 10 is highlighted in blue. Notably, Whitendale (the site with some previous mowing in both catchments) did not show any consistent impact whereas Nidderdale and Mossdale both showed an about 15% lower loss in mown compared to burnt catchments. However, this difference in flow volume does not relate to burning impacts directly, as it does not compare burning to no management, it rather compares mowing versus burning, highlighting that mowing with leaving brash can increase the overall water holding capacity of a bog with potential implications on reducing flooding downstream (although this remains to be tested and requires further analysis of peak flow volumes and storm flow behaviour and possible effects on evaporation rates).