Modeling forest management effects on water and sediment yield from nested, paired watersheds in the interior Pacific Northwest, USA using WEPP

WEPPcloud hydrologic and erosion simulation datasets from 28 watersheds in US Pacific Northwest and calibrating model parameters for undisturbed and disturbed forest management conditions

The WEPPcloud interface is a new online decision-support tool for the Water Erosion Prediction Project (WEPP) model that facilitates data preparation and model runs, and summarizes model outputs into tables and maps that are easily interpretable by users. The interface can be used by land and water managers in United States, Europe, and Australia interested in simulating streamflow, sediment and pollutant loads from both undisturbed and disturbed (e.g. post-wildfire or post-treatment such as thinning or prescribed fires) forested watersheds. This article contains full hydrologic model runs for 28 forested watersheds in the U.S. Pacific Northwest with the WEPPcloud online interface. It also includes links to repositories with the individual model runs, a table containing default model parameters for disturbed conditions, and figures with model outputs as compared to observed data. The data in the repositories include all the raw data input and output from the model as well as the processed data, which can be accessed through tables and shapefiles to provide additional insights into the model outputs. Lastly, the article describes how the data are organized and the content of each folder containing the data. These model runs are useful for anyone interested in modeling forested watersheds with the WEPPcloud interface.

Optimizing invasive species management using mathematical programming to support stewardship of water and carbon-based ecosystem services

Invasive species alter hydrologic processes at watershed scales, with impacts to biodiversity and the supporting ecosystem services. This effect is aggravated by climate change. Here, we integrated modelled hydrologic data, remote sensing products, climate data, and linear mixed integer optimization (MIP) to identify stewardship actions across space and time that can reduce the impact of invasive species. The study area is the windward coast of Hawai'i Island (USA) across which non-native strawberry guava occurrence varies from extremely dense stands in lower watershed reaches, to low densities in upper watershed forests. We focused on the removal of strawberry guava, an invader that exerts significant impacts on watershed condition. MIP analyses spatially optimized the assignment of effective management actions to increase water yield, generate revenue from enhanced freshwater services, and income from removed biomass. The hydrological benefit of removing guava, often marginal when considered in isolation, was financially quantified, and single- and multiobjective MIP formulations were then developed over a 10-year planning horizon. Optimization resulted in $2.27 million USD benefit over the planning horizon using a payment-for-ecosystem-services scheme. That value jumped to $4.67 million when allowing work schedules with overnight camping to reduce costs. Pareto frontiers of weighted pairs of management goals showed the benefit of clustering treatments over space and time to improve financial efficiency. Values of improved land-water natural capital using payment-for-ecosystem-services schemes are provided for several combinations of spatial, temporal, economical, and ecosystem services flows.

Improving and calibrating channel erosion simulation in the Water Erosion Prediction Project (WEPP) model

The Water Erosion Prediction Project (WEPP) model has been widely used to assess the impacts of management practices and climate change on runoff and soil loss at both hillslope and watershed scales. However, the representation of channel erosion processes in WEPP has not been changed significantly since it was released. The current (WEPP v2012.8) and previous WEPP versions assume that channel input erodibility parameters are constant through time, which may lead to erroneous channel detachment predictions, especially for cropland with substantial tillage operations. In this research, the temporally constant values of channel erodibility and critical shear stress were replaced by daily updated values, using the same temporal erodibility and critical shear stress adjustments that are applied in hillslope profile simulations for rill detachment. Observed watershed-scale runoff and soil erosion data from six agricultural watersheds were used to calibrate and compare the WEPP model performance in simulating channel runoff volumes and soil losses before and after the modification. The research showed both WEPP v2012.8 and the modified WEPP model (WEPP_CE) could satisfactorily simulate event-based hydrology and soil erosion at the watershed outlets after calibration. The WEPP_CE model with temporally varying channel erodibility and critical shear stress values demonstrated improved representation of the physical processes in channel soil detachment. Continued improvement in the representation of channel erosion processes in WEPP and other process-based models is needed. The improved WEPP model can be used to evaluate the effectiveness of soil conservation practices on hydrology and erosion in further research.

Do Land Use Changes Balance out Sediment Yields under Climate Change Predictions on the Sub-Basin Scale? The Carpathian Basin as an Example

The issue of whether land use changes will balance out sediment yields induced by climate predictions was assessed for a Carpathian basin (Raba River, Poland). This discussion was based on the Macromodel DNS (Discharge–Nutrient–Sea)/SWAT (Soil and Water Assessment Tool) results for the RCP 4.5 and RCP 8.5 scenarios and LU predictions. To track sediment yield responses on the sub-basin level the studied area was divided into 36 units. The response of individual sub-basins to climate scenarios created a mosaic of negative and positive sediment yield changes in comparison to the baseline scenario. Then, overlapped forest and agricultural areas change indicated those sub-basins where sediment yields could be balanced out or not. The model revealed that sediment yields could be altered even by 49% in the selected upper sub-basins during the spring-summer months, while for the lower sub-basins the predicted changes will be less effective (3% on average). Moreover, the winter period, which needs to be re-defined due to an exceptional occurrence of frost and snow cover protecting soils against erosion, will significantly alter the soil particle transfer among the seasons. Finally, it has been shown that modeling of sediment transport, based on averaged meteorological values and LU changes, can lead to significant errors.

Using Machine Learning-Based Algorithms to Analyze Erosion Rates of a Watershed in Northern Taiwan

This study continues a previous study with further analysis of watershed-scale erosion pin measurements. Three machine learning (ML) algorithms—Support Vector Machine (SVM), Adaptive Neuro-Fuzzy Inference System (ANFIS), and Artificial Neural Network (ANN)—were used to analyze depth of erosion of a watershed (Shihmen reservoir) in northern Taiwan. In addition to three previously used statistical indexes (Mean Absolute Error, Root Mean Square of Error, and R-squared), Nash–Sutcliffe Efficiency (NSE) was calculated to compare the predictive performances of the three models. To see if there was a statistical difference between the three models, the Wilcoxon signed-rank test was used. The research utilized 14 environmental attributes as the input predictors of the ML algorithms. They are distance to river, distance to road, type of slope, sub-watershed, slope direction, elevation, slope class, rainfall, epoch, lithology, and the amount of organic content, clay, sand, and silt in the soil. Additionally, measurements of a total of 550 erosion pins installed on 55 slopes were used as the target variable of the model prediction. The dataset was divided into a training set (70%) and a testing set (30%) using the stratified random sampling with sub-watershed as the stratification variable. The results showed that the ANFIS model outperforms the other two algorithms in predicting the erosion rates of the study area. The average RMSE of the test data is 2.05 mm/yr for ANFIS, compared to 2.36 mm/yr and 2.61 mm/yr for ANN and SVM, respectively. Finally, the results of this study (ANN, ANFIS, and SVM) were compared with the previous study (Random Forest, Decision Tree, and multiple regression). It was found that Random Forest remains the best predictive model, and ANFIS is the second-best among the six ML algorithms.