ADVANCING HYDROSTRATIGRAPHIC MODEL (HSM) AND GROUNDWATER NUMERICAL MODEL DEVELOPMENT WITH AEM AND HYDROGEOLOGIC DATA
The toolset – Data2HSM, Data2Texture, and Texture2Par - facilitates the processing, integration, and codification of disparate data types for water resource planning and modeling. These tools can function as stand-alone products, each incorporating different datasets for the user’s end goals, or work together to generate 3D hydrostratigraphic models (HSMs) or numerical models.
Data2HSM is a suite of three methods that utilize machine learning algorithms to process AEM and other hydrogeologic data, streamlining the creation of HSMs. The Gaussian Mixture Model clusters hydrostratigraphic units from AEM data, Smart Interpretation automatically interprets geologic surfaces from user-selected training points, and GeoPDNN models stratigraphic surfaces using point-based datasets.
Data2Texture is an advanced spatial data interpolation tool for estimating the distribution of sediment textures using AEM and boring log data. The application of Data2Texture includes the development of 3D texture models.
Texture2Par is a groundwater model pre-processor and parameterization utility developed to work with IWFM and MODFLOW codes. Texture2Par takes the interpolated, gridded texture data from Data2Texture and estimates aquifer and aquitard hydraulic parameters (hydraulic conductivity, specific yield, specific storage) for groundwater models or other applications.
This presentation shares progress updates on the toolset and seeks input on development directions to ensure the final toolset offers the most beneficial capabilities to the groundwater community.