Repository logo

Browsing by Author "Bhaskar, Aditi S., committee member"

Now showing 1 - 2 of 2
  • Thumbnail Image
    ItemOpen Access
    Multiscale connections between a groundwater dependent ecosystem and socio-hydrology: insight gained from numerical modeling, geospatial informatics, and Bayesian statistics
    (Colorado State University. Libraries, 2023) Lurtz, Matthew R., author; Morrison, Ryan R., advisor; Bhaskar, Aditi S., committee member; Bailey, Ryan T., committee member; Ross, Matthew, committee member
    The connectivity between floodplain practices and groundwater dependent ecosystems (GDE) is undeniable, yet difficult to measure. Quantifying the connection between ecosystems would be ideal for the conjunctive management of groundwater and surface water resources in an irrigated river valley. In the research presented, a variety of methodologies are used to understand the socio-hydrologic connections between a semi-arid GDE and agro-pastoral practices in southeastern Colorado (USA). I investigated the socio-hydrologic relationships between a GDE and the surrounding floodplain using three approaches. First, I used the output from a calibrated groundwater model and a remote sensing evapotranspiration (ET) algorithm with exploratory statistics. Second, I used remotely sensed vegetation information and socio-hydrologic data in a Bayesian hierarchical time series and spatial statistics models to compliment the first approach by examining new explanatory covariates. Third, a simple regression framework examines the point-scale relationship between groundwater and ET to further dissect results from the first approach at a finer resolution. These three approaches yielded key results. From the first objective, the dual-model comparison agreed with previous ecological research showing a non-linear relationship between ET and groundwater depth (0-5 m), and a threshold was identified at three meters where the rate between ET and groundwater depth change. The time series and spatial statistics objective helped identify a spatial scale threshold to detect temporal trend, lagged intra-seasonal predictors of vegetation water use, and which floodplain characteristics impact vegetation density. This statistical analysis discovered that temporal trend is not detectable at spatial scales larger than catchment size (> 10 km). Monthly temperature and lagged monthly values of precipitation and stream gain-loss (i.e., an return flow indicator variable) are all predictive of temporal changes in riparian vegetation density. Based on the floodplain characteristics tested in the spatial statistics approach, perennial tributaries to the Arkansas River increase vegetation density while the conversion of agriculture to fallow land decrease riparian vegetation density. The third objective highlighted that the process between evapotranspiration and groundwater head is non-linear but depends on temporal scale and plant functional group. The results from these approaches is important for GDE preservation in the face of increasing demand on groundwater supply. The process between groundwater and ET is of particular importance in large scale water balance studies that include a groundwater and surface water interface with need to model the groundwater-ET relationship in natural and agricultural ecosystems simultaneously.
  • Thumbnail Image
    ItemOpen Access
    Towards interactive betweenness centrality estimation for transportation network using capsule network
    (Colorado State University. Libraries, 2022) Matin, Abdul, author; Pallickara, Sangmi Lee, advisor; Pallickara, Shrideep, committee member; Bhaskar, Aditi S., committee member
    The node importance of a graph needs to be estimated for many graph-based applications. One of the most popular metrics for measuring node importance is betweenness centrality, which measures the amount of influence a node has over the flow of information in a graph. However, the computation complexity of calculating betweenness centrality is extremely high with large- scale graphs. This is especially true when analyzing the road networks of states with millions of nodes and edges, making it infeasible to calculate their betweenness centrality (BC) in real- time using traditional iterative methods. The application of a machine learning model to predict the importance of nodes provides opportunities to address this issue. Graph Neural Networks (GNNs), which have been gaining popularity in recent years, are particularly well-suited for graph analysis. In this study, we propose a deep learning architecture RoadCaps to estimate the BC by merging Capsule Neural Networks with Graph Convolutional Networks (GCN), a convolution operation based GNN. We target the effective aggregation of features from neighbor nodes to approximate the correct BC of a node. We leverage patterns capturing the strength of the capsule network to effectively estimate the node level BC from the high-level information generated by the GCN block. We further compare the model accuracy and effectiveness of RoadCaps with the other two GCN-based models. We also analyze the efficiency and effectiveness of RoadCaps for different aspects like scalability and robustness. We perform one empirical benchmark with the road network for the entire state of California. The overall analysis shows that our proposed network can provide more accurate road importance estimation, which is helpful for rapid response planning such as evacuation during wildfires and flooding.