Use cases

Visualizing time series at two different sites is a simple and easy way to explore differences among those sites. In the case of restoration, we often want to compare temporal dynamics at the restoration site and a “control”, i.e, a site that is similar to the restoration site, but has not had any restoration activity.

This use case describes the workflow for delineating sites in a MRRMaid app, downloading the time-series data and plotting the data from the two sites on a single graph, using Excel software.

Video overview transcript here.

In observational studies, it can be challenging to attribute cause to a specific action, e.g. did restoration treatments cause an increase in floodplain vegetation? Luckily, we can use quasi-experimental methods that can answer these types of questions making use of the 40+ years of Landsat satellite data available.

This use case uses quasi-experimental analysis, the MRRMaid MVPRestore data, and drought data from Climate Engine to estimate (1) the amount and direction of change in floodplain vegetation and (2) the probability that the change was caused by the restoration treatments.

Video overview transcript for the second use case here.

The main assumption of the MRRMaid Toolbox is that vegetation extent is a good proxy for water availability. The strength and validity of this assumption can be tested in watersheds where long-term streamflow data is available.

This use case uses correlation analyses to quantify the relationships between time series of floodplain vegetation, hydrology, and climate at the decadal time scale.

Video overview transcript for the third use case here.

Prioritization of restoration activities requires comparison of systems across space. To facilitate these comparisons, we chose to summarize mesic vegetation in valley bottoms within each management unit, excluding irrigated agriculture. We use two simple metrics, mean and variability, to describe the mesic ecosystems of each unit.

The workflow we outline here provides information about how we determined the proportion and the variability of mesic vegetation estimated in the valley bottoms of every HUC12 unit in Owyhee County.

This use case describes the methods we outline in the manuscript titled Time series analysis to demonstrate restoration outcomes and system change from satellite data, published in Restoration Ecology. While the manuscript goes into detail about technical aspects of each method, the goal of this document is to demonstrate how any user can apply these time series analysis methods to their own site using the Monthly MRRMaid time series.

Ecologists are increasingly utilizing observational data sources to assess ecological outcomes. However, traditional experimental designs do not apply with these data because there is no appropriate set of ‘control’ observations to evaluate treatment. Economists, however, have developed several impact evaluation methods that are gradually becoming more prevalent in ecological studies.

In this use case, we introduce both counterfactual matching and difference in differences regression as a way to estimate causal impact of a given treatment using an observational outcome dataset.

Video transcript for the sixth use case here.