Journal Paper Digests 2023 #11
- An integrated approach for estimating soil health: Incorporating digital elevation models and remote sensing of vegetation
- Mapping Landslide Susceptibility Over Large Regions With Limited Data
- An Unsaturated Hydraulic Conductivity Model Based on the Capillary Bundle Model, the Brooks-Corey Model and Waxman-Smits Model
- Conceptualising the drivers of ultra-processed food production and consumption and their environmental impacts: A group model-building exercise
Conceptualising the drivers of ultra-processed food production and consumption and their environmental impacts: A group model-building exercise
Using group model building we developed a series of causal loop diagrams identifying the environmental impacts of ultra-processed food (UPF) systems, and underlying system drivers, which was subsequently validated against the peer-reviewed literature. The final conceptual model displays the commercial, biological and social drivers of the UPF system, and the impacts on environmental sub-systems including climate, land, water and waste. It displays complex interactions between various environmental impacts, demonstrating how changes to one component of the system could have flow-on effects on other components. Trade-offs and uncertainties are discussed. The model has a wide range of applications including informing the design of quantitative analyses, identifying research gaps and potential policy trade-offs resulting from a reduction of ultra-processed food production and consumption.
An Unsaturated Hydraulic Conductivity Model Based on the Capillary Bundle Model, the Brooks-Corey Model and Waxman-Smits Model
Soil unsaturated hydraulic conductivity (K), which depends on water content (?) and matric potential (?), exhibits a high degree of variability at the field scale. Here we first develop a theoretical hydraulic-electrical conductivity (s) relationship under low and high salinity cases based on the capillary bundle model and Waxman and Smits model which can account for the non-linear behavior of s at low salinities. Then the K-s relationship is converted into a K(?, ?) model using the Brooks-Corey model. The model includes two parameters c and ?. Parameter c accounts for the variation of the term (? + 2)/(? + 4) where ? is the pore size distribution parameter in the Brooks-Corey model, and the term m-n where m and n are Archie’s saturation and cementation exponents, respectively. Parameter ? is the sum of the tortuosity factor accounting for the differences between hydraulic and electrical tortuosity and Archie’s saturation exponent. Based on a calibration data set of 150 soils selected from the UNSODA database, the best fitting log(c) and ? values were determined as -2.53 and 1.92, -4.39 and -0.14, -5.01 and -1.34, and -5.79 and -2.27 for four textural groups. The estimated log(10)(K) values with the new K(?, ?) model compared well to the measured values from an independent data set of 49 soils selected from the UNSODA database, with mean error (ME), relative error (RE), root mean square error (RMSE) and coefficient of determination (R-2) values of 0.02, 8.8%, 0.80 and 0.73, respectively. A second test of the new K(?, ?) model using a data set representing 23 soils reported in the literature also showed good agreement between estimated and measured log(10)(K) values with ME of -0.01, RE of 9.5%, RMSE of 0.77 and R-2 of 0.85. The new K(?, ?) model outperformed the Mualem-van Genuchten model and two recently published pedo-transfer functions. The new K(?, ?) model can be applied for estimating K under field conditions and for hydrologic modeling without need for soil water retention curve data fitting to derive a K function.
Mapping Landslide Susceptibility Over Large Regions With Limited Data
Landslide susceptibility maps indicate the spatial distribution of landslide likelihood. Modeling susceptibility over large or diverse terrains remains a challenge due to the sparsity of landslide data (mapped extent of known landslides) and the variability in triggering conditions. Several different data sampling strategies of landslide locations used to train a susceptibility model are used to mitigate this challenge. However, to our knowledge, no study has systematically evaluated how different sampling strategies alter a model’s predictor effects (i.e., how a predictor value influences the susceptibility output) critical to explaining differences in model outputs. Here, we introduce a statistical framework that examines the variation in predictor effects and the model accuracy (measured using receiver operator characteristics) to highlight why certain sampling strategies are more effective than others. Specifically, we apply our framework to an array of logistic regression models trained on landslide inventories collected at sub-regional scales over four terrains across the United States. Results show significant variations in predictor effects depending on the inventory used to train the models. The inconsistent predictor effects cause low accuracies when testing models on inventories outside the domain of the training data. Grouping test and training sets according to physiographic and ecological characteristics, which are thought to share similar triggering mechanisms, does not improve model accuracy. We also show that using limited landslide data distributed uniformly over the entire modeling domain is better than using dense but spatially isolated data to train a model for applications over large regions.
An integrated approach for estimating soil health: Incorporating digital elevation models and remote sensing of vegetation
The deterioration of soil health (SH) in agricultural lands is a global challenge that poses a threat to food and resource security. We developed a practical framework to facilitate the large-scale SH assessment in agricultural fields of northwestern Iran. A total of 350 soil samples were collected and soil properties were determined. Eight linear and non-linear Soil Health Indexes (SHIs) were developed. Digital Elevation Model (DEM) and multiple remote sensing indexes were obtained from satellite images. SHI prediction models were developed using an integrated approach and through a model selection procedure, the most relevant indexes were identified. The results showed significant (P < 0.05) positive correlation between the IHI-LT and elevation (r = 0.56), Vegetation Health Index (VHI) (r = 0.69), and Surface Water Condition Index (SWCI) (r = 0.79). The multiple regression model including the above indexes strongly explained the spatial variability of the Integrated Soil Health Index (IHI) with both total (LT) and minimum (LM) dataset approaches (R2 = 0.72; AIC =-1607.27; RMSE = 0.03; rho c = 0.65). The developed models can be utilized for large-scale assessment of soil health conditions, reducing the cost and effort of conventional ground-truth soil sampling and analysis. Furthermore, this approach may aid in monitoring and mitigating the soil degradation in agricultural lands.