Evaluating spatially enabled machine learning approaches to depth to bedrock mapping, Alberta, Canada.

Maps showing the thickness of sediments above the bedrock (depth to bedrock, or DTB) are important for many geoscience studies and are necessary for many hydrogeological, engineering, mining, and forestry applications. However, it can be difficult to accurately estimate DTB in areas with varied topography, like lowland and mountainous terrain, because traditional methods of predicting bedrock elevation often underestimate or overestimate the elevation in rugged or incised terrain. Here, we describe a machine learning spatial prediction approach that uses information from traditional digital elevation model derived estimates of terrain morphometry and satellite imagery, augmented with spatial feature engineering techniques to predict DTB across Alberta, Canada. First, compiled measurements of DTB from borehole lithologs were used to train a natural language model to predict bedrock depth across all available lithologs, significantly increasing the dataset size. The combined data were then used for DTB modelling employing several algorithms (XGBoost, Random forests, and Cubist) and spatial feature engineering techniques, using a combination of geographic coordinates, proximity measures, neighbouring points, and spatially lagged DTB estimates. Finally, the results were contrasted with DTB predictions based on modelled relationships with the auxiliary variables, as well as conventional spatial interpolations using inverse-distance weighting and ordinary kriging methods. The results show that the use of spatially lagged variables to incorporate information from the spatial structure of the training data significantly improves predictive performance compared to using auxiliary predictors and/or geographic coordinates alone. Furthermore, unlike some of the other tested methods such as using neighbouring point locations directly as features, spatially lagged variables did not generate spurious spatial artifacts in the predicted raster maps. The proposed method is demonstrated to produce reliable results in several distinct physiographic sub-regions with contrasting terrain types, as well as at the provincial scale, indicating its broad suitability for DTB mapping in general.

Prevalence and progression of macroscopic lesions in Orbicella annularis and O. faveolata on shallow fringing reefs of St. Kitts.

The endangered corals Orbicella annularis and O. faveolata are crucial to Caribbean reefs because of their large size and contribution to reef framework. The objective of this study was to describe the prevalence and progression of macroscopically evident lesions affecting Orbicella spp. in shallow fringing reefs in St. Kitts. Cross-sectional surveys in the spring of 2017 demonstrated 8 predominant lesion patterns affecting 59% of corals (95% CI: 55.8-62.1%), including annular yellow-brown pigmentation, focal brown pigmentation, focal bleaching, diffuse bleaching, annular black surface deposit, focal tissue loss with skeletal erosion, focal grey pigmentation, and growth anomaly. Longitudinal surveys of 47 tagged corals were performed from August 2016-May 2017 to track lesion progression. The 2 most common lesions, annular yellow-brown pigmentation (n = 30), and focal brown pigmentation (n = 21), showed mean (±SD) partial colony mortality growth of 0.26 ± 0.5 and 0.21 ± 0.45 cm2 d-1, respectively. Annular pigmentation progression severity was associated with a marginating band of bleaching (ordinal odds ratio [OOR] = 11.0), and yellow rather than brown color (OOR = 3.8). Bleaching lesions (n = 13), occurring during a time of elevated sea surface temperature, were most severe during October-December 2016, and persisted through April 2017, months after heat stress had subsided. Annular black surface deposits (n = 3) were associated with rapid progression of acute tissue loss, whereas focal tissue loss with skeletal erosion (n = 2) regressed within months, and focal grey pigmentation (n = 2) was quiescent for the length of the study. This study enforces concern for the extent to which Orbicella spp. are declining due to disease.