Field Monitoring Techniques
For watersheds, slopes, and river reaches that rated moderate risk for the RRISSC procedure, the implementation of process specific mitigation or management prescriptions needs to be monitored. Monitoring will determine the appropriateness or effectiveness of specific management prescriptions designed and implemented to reduce potential adverse sediment and/or river stability effects. Since monitoring requires site specific measurements, parameters that reflect temporal, spatial, scale, streamflow variation, and site/reach representation are required in order to properly represent situational variability and extrapolate findings of a process and/or system response to imposed change. Such variability factors are summarized as
To isolate the variability of season and/or annual change, monitoring design should include time scales. For example, measuring annual lateral erosion rates should include measurements once per year at the same time of year. If the objectives are to identify seasons where disproportionate erosion occurs, measurements may be obtained during snowmelt runoff, later post stormflow runoff, ice-off, and other periods of time associated with a given erosional process. For example, annual replicate surveys of particle size gradation of bed material under a permanent cross-section of a glide will provide valuable information of magnitude, direction, and consequence of annual shifts.
Temporal measurements must also cover a range of time during bedload sampling because surges or slugs of bedload often appear. Sampling over recommended time periods for a given flow (generally 20 minutes) improves the probability of observing this variability (as opposed to an instantaneous point sample). Short-term versus long-term monitoring must also be considered based on the probability of change, the severity and consequence of effects, and the likelihood of variation. Sampling over many years, although costly, may be warranted to cover changes in wet/dry periods, etc.
Variability of change/response involving spatial considerations can be identified by measurements of the same process at more than one site (cross-section) or by more intense monitoring at the same site. For example, a longitudinal profile measured over a couple of meander wavelengths will indicate changes in the maximum depth and/or slope of pools, which cannot be derived bymonitoring one pool at one location. Identifying more than one reach of the same morphological type can also be used to understand response trends. Sampling the spatial variability (both vertically and laterally) within a cross-section of velocity and sediment types helps identify or at least integrate such variability into a documented observation.
Monitoring streams of various sizes and or stream orders, but of the same morphological type and condition, will help identify variability in system response for proper extrapolation of results. For example, vertical stability measurements should be made on river reaches of the same condition and the same type but at locations that reflect various stream widths (size) and stream order.
Measurements of channel process relationships need to be stratified over a range of seasonal and annual flows. For example, both suspended and bedload sediment should be measured over a wide range of flows during the freshet, low-elevation snowmelt, high elevation snowmelt, rising versus recession stages, stormflow runoff, and baseflow. This stratification for streamflow allows the field observer to plot a sediment rating curve that represents the widest range of seasonal flows within whichchanges in sediment supply can vary.
Site or reach variation
Monitoring a site for soil loss should include a soil type designation for potential extrapolation for similar conditions on similar soil types. The same is true for stream types. Sediment, hydraulic, and stability monitoring needs to be stratified by stream type since such data will naturally vary for the reference (stable) reach between stream types. This information is helpful indetecting departure from a reference stream type rather than differences among stream types.
Often the best laid plans are not implemented due to various reasons. Characterizing the response of a process and/or system to management actions must first address whether or not mitigation was implemented correctly. Riparian vegetation response due to livestock exclusion with riparian fencing would be very misleading if fence maintenance failed or gates were left open, etc. If restoration designs were correct but the contractor installed structures at the wrong angle, slope, or position on the bank, then near-bank stress reduction or erosion rate would likely not be a fair observation related to the effectiveness of the mitigation structure.
Without proper stratification and design of monitoring for measurements sensitive to temporal, spatial, scale, streamflow change, soil and river types and problems with implementation, it is difficult to determine departure or to identify geologic from anthropogenic sources. For example, one sediment data point collected at the mouth of a large order drainage has little value in specifying the source, natural versus anthropogenic influence, or identifying process or land use contributions. Past monitoring efforts have been less than effective in many cases, often due to measurements that do not explain variability in natural systems or inconsistencies due to mitigation implementation problems.