A large proportion of the uncertainty surrounding catchment sediment budget modelling has been attributed to sediment supplied from riverbank erosion. Some of the variables influencing riverbank erosion are bend curvature, specific streampower, riparian vegetation, and in some instances sand and gravel extraction. The empirical relationship between these variables and observed riverbank erosion across 78 km of the Upper Brisbane River, Australia was investigated. No significant relationship was found between curvature, specific streampower and riverbank erosion. The role of riparian vegetation relative to sediment supply from riverbank erosion varied with spatial location, susceptibility of a reach to erosion, and human disturbance such as sand and gravel extraction. Despite not having data on substrate type the model described approximately 37% of the variation in observed riverbank erosion. It appears that inclusion of a management practice factor in riverbank erosion models is justified, where appropriate, and may improve model performance.Griffith Sciences, Griffith School of EnvironmentFull Tex
Livestock grazing is a major land use in the Great Barrier Reef Catchment Area (GBRCA). Heightened grazing density coupled with inadequate land management leads to accelerated soil erosion and increased sediment loads being transported downstream. Ultimately, these increased sediment loads impact the water quality of the Great Barrier Reef (GBR) lagoon. Ground cover mapping has been adopted to monitor and assess the land condition in the GBRCA. However, accurate prediction of ground cover remains a vital knowledge gap to inform proactive approaches for improving land conditions. Herein, we explored two deep learning-based spatio-temporal prediction models, including convolutional LSTM (ConvLSTM) and Predictive Recurrent Neural Network (PredRNN), to predict future ground cover. The two models were evaluated on different spatial scales, ranging from a small site (i.e., <5 km2) to the entire GBRCA, with different quantities of training data. Following comparisons against 25% withheld testing data, we found the following: (1) both ConvLSTM and PredRNN accurately predicted the next-season ground cover for not only a single site but also the entire GBRCA. They achieved this with a Mean Absolute Error (MAE) under 5% and a Structural Similarity Index Measure (SSIM) exceeding 0.65; (2) PredRNN superseded ConvLSTM by providing more accurate next-season predictions with better training efficiency; (3) The accuracy of PredRNN varies seasonally and spatially, with lower accuracy observed for low ground cover, which is underestimated. The models assessed in this study can serve as an early-alert tool to produce high-accuracy and high-resolution ground cover prediction one season earlier than observation for the entire GBRCA, which enables local authorities and grazing property owners to take preventive measures to improve land conditions. This study also offers a new perspective on the future utilization of predictive spatio-temporal models, particularly over large spatial scales and across varying environmental sites.
If soil resources and the benefits derived from water quality are to be maintained, the on- and off-site effects of soil erosion must be adequately represented and communicated to decision makers so that appropriate management responses can be identified. The System of Environmental-Economic Accounting - Ecosystem Accounting (SEEA-EA), is one approach to quantify both the contributions that ecosystems make to the economy, and the impacts of economic activity on ecosystems. However, due to the difficulty of obtaining empirical data on ecosystem service flows, in many cases such quantification is informed by ecosystem service models. Previous research on the Mitchell catchment, Queensland Australia allowed us to explore the implications of using a model of hillslope erosion and sediment delivery in isolation (as represented in one of the most frequently used ecosystem service models - InVEST) by comparing such estimates against multiple lines of local empirical data, and a more comprehensive representation of locally important erosion and deposition processes through a sediment budget model. Estimates of the magnitude of hillslope erosion modelled using an approach similar to InVEST and the calibrated sediment budget differed by an order of magnitude. If an uncalibrated InVEST-type model was used to inform the relative distribution of erosion magnitude, findings suggest the incorrect erosion process would be identified as the dominant contributor to suspended sediment loads. However, the sediment budget model could only be calibrated using data on sediment sources and sinks that had been collected in the catchment through a sustained research effort. A comparable level of research investment may not be available to inform ecosystem service assessments elsewhere. The results summarised here for the Mitchell catchment demonstrate that practitioners must exercise caution when using estimates of the sediment retention ecosystem service flow which have not been calibrated and validated against locally collected empirical data.