We have developed an applied case of automatic calibration with PEST over a MODFLOW 6 model for a pumping test. The model represents a 72 hour pumping test over a confined aquifer with an observation point located 11 meters from the well.  Model Muse was used for the model construction, Pest activation and parameter setup.

We have developed an applied case of automatic calibration with PEST over a MODFLOW 6 model for a pumping test. The model represents a 72 hour pumping test over a confined aquifer with an observation point located 11 meters from the well.  Model Muse was used for the model construction, Pest activation and parameter setup.

This webinar covers the procedure to implement a neural network based on a set of parameters set with corresponding head values; the study case is on a transient pumping test model with an observation point located around 10 meters away from the pump well. The analysis predicts the calibration parameters from the spatially interpolated observed head values. 

Sensitivity analysis is referred to the uncertainty analysis in model results from the uncertainties on the model inputs (or something close to this definition). We usually perform this analysis on observation points (point that we know) with certain predefined ranges over the hydraulic parameters.

Python is an object oriented programming language and everything can be treated as an object like a MODFLOW-6 model. Having a pumping test model as an object allows us to use powerful Python libraries as SALib to perform sensibility analysis. 

This webinar covers the whole procedure to perform a sensitivity analysis over a 72 hour pumping test plus recovery on the hydraulic response in an observation piezometer located at 11 meters from the well. the Since the study case it’s a transient model the sensitivities will vary over time and stage of pumping/recovery.

We have done an applied groundwater model for a pumping test on a confined aquifer. The wellsite is located on the Clairborne aquifer (Georgia, USA) and has two observations on different layers besides the well itself. The aquifer test has two periods of pumping and recovery that last 3 days each one and the numerical model was constructed with MODFLOW-6 based on a Python script with the FloPy package. Comparison among observed and simulated wells were done with plots in Matplotlib.

An applied case for the recognition of missing crop vegetation areas based on a drone orthophoto. Contours have been identified from an enhanced combination of raster bands with a marching squares method to find constant valued contours and then exported as geospatial polygons.

NetCDF has become a popular choice for storing and delivering precipitation and water resources related data. Its capacities to store multiple geospatial raster layers over time allow another level of abstraction on data analysis, however the format is of limited use on normal desktop applications and most times we are required to use a programming language such as Python or R.

We have done an applied example of exploration analysis from a NetCDF file that has the CHIRPS dataset for year 2022. The dataset is opened with Python and Rasterio and clipped to country extent where an interactive visualization of precipitation is performed. Finally, a daily precipitation hydrograph is generated over the coordinates of a given city.

If you want to speed some steps in groundwater modeling you might be interested in the tools presented on this webinar. We have developed some online tools that replace intense and time consuming tasks on desktop software and web servers and provide vector and raster data on the formats required by Model Muse. The webinar covers the steps of DEM generation as Surfer Grid File (*.grd) with the definition of watersheds and river network based on two online services provided by Hatarilabs.

Modeling tidal waves on groundwater flow models is a challenge due to the several stress periods required to represent the wave and the response on the aquifer system. Standard tools on GUI software fall short to implement multiple stress periods, multiple boundary condition parameters and represent selected output with the tidal elevations.

We have done an “hybrid” approach where a coastal aquifer is built on Model Muse and later imported by Flopy where the boundary conditions that represent the sea elevation is replaced by a time series, new stress periods are inserted, workspace has changed and the model is simulated. The applied case ends up with the head representation and the comparison among the tidal elevation and the piezometers located at different distances and depths.

Large amount of spatial data is indexed and delivered through files in Hierarchical Data Format (HDF). These files are compatible with desktop GIS software as QGIS but they are not so easy to open/read/process with standard Python libraries as Rasterio, or with dedicated libraries. On our research we found the spatial functionality on the powerful GDAL binaries and library for Python. 

We have done an applied case for the download, reproject (to WGS84) and clip to the area of interest of several MYD16A2 actual evapotranspiration images. We have also create an interactive visualization of the clipped images that allow us to examine the precipitation trends and the challenges in the use of these data products. 

Having more tools and managing more data enhances us to reach a superior level of data analysis. This time we have developed an applied case of automatic calibration with PEST of a previously manual calibrated MF6 model on a hillslope area. Our main purpose was to analyze the increase / decrease in the calibration statistics for the automatic versus the manual calibration model and discuss what the new parameters tell us from the original hydrogeological conceptual model. 

Modeling flow on open channels and simulating water mixing to evaluate the final concentrations and mixing area was a hard task with conventional open source software. Based on our research and app development we came up with a process to model flow of water and effluent mixing and then calculate the mass flux with the initial concentration of certain chemical components to have a fully 3d simulation of the concentration development with time and distance from source. 

As for the myth of Icarus who wanted to fly close to the sun, sometimes normal Qgis users want to fly close to the web however our “wings” are unknown. Qgis2Web is a robust Qgis plugin that is 8 years old, has more than 177 releases and allows a normal Qgis user to fly to the web with minimal knowledge of html, javascript or other web standards. 

The processes of working with digital elevation models (DEMs) and watershed delineation are in constant change over time. Even though the elevation datasets date from 10 years ago; the web servers and tools for data processing have changed in recent years. With online tools such as Hatari Utils, the watershed delineation can be performed in a few steps with practical outputs as vector files and watershed statistics.

This webinar covers the whole process of defining an area of interest in QGIS, export it to NASA Earthdata Search, review the elevation datasets available, download multiple elevation rasters, join rasters and clip to the area of interest and delineate the watershed online.

If you want to apply a complete 3D geological model into a groundwater flow model, this is the webinar that you were looking for. Geological models generated with Python and Gempy and stored as voxels (x,y,z,lito) can be imported into a Modflow model with Flopy. This webinar covers an applied example of geological model import, massive spatial queries to determine the corresponding cell (layer, row, col), lithological filtering and hydraulic parameter determination based on lithology into a Modflow model. Results can be imported into Model Muse for further model development, boundary condition setup and simulation.

The applied example comprises the translation of a geological model of 196875 voxels (35 layers, 75 rows, 75 cols) into a MODFLOW model of 18000 cells (20 layers, 30 rows, 30 cols).

The distribution of geological units over different depths on an area of interest is often needed as an input for other numerical models. Sources of data for geological modeling are usually bore logs but these can be also interpreted geological unit bottom elevations as geospatial rasters. Based on a coupled workflow on QGIS and Python it is possible to extract the required information for a Gempy model and run it for defined voxel sizes. This webinar covers the whole procedure of spatial data preparation, data preprocessing in defined formats and geological modeling with Python and Gempy.

Gempy is an open-source library for modeling geology written in Python. The library is capable of creating complex 3D geological models including structures, fault networks, and unconformities and it can be coupled with uncertainty analysis.

This webinar develops an applied case of a 5 layer geological model with a fault developed from information of geological contacts and surface orientations. The webinar covers from the steps of Gempy installation in Windows, the geological model setup and simulation, to the geological model export as VTK.

This webinar covers an applied case of a numerical model model review from the report and published open data for the Rustler Aquifer, Texas, USA. The review is done with scripts of Flopy that represent the report figures and open the output data to extract water heads on selected wells.

Groundwater model creation requires a complete set of spatial data for the different hydraulic parameters, boundary conditions and other model items. Vector and raster data need to be preprocessed, converted, reprojected to fit the requirements of Model Muse.

This webinar will cover an applied case of raster and vector data processing for a groundwater model over the Oahu island (Hawaii, USA). The webinar covers the steps to download the data, extract the river network and import the spatial data into Model Muse.

This webinar covers the complete procedure to create a land cover change raster from a comparison of generated vegetation index (NDVI) rasters by the use of Python and the Numpy and Rasterio libraries. Results of the NDVI for given years and NDVI change are plotted on Jupyter Lab as color grid and contour grid. An analysis of deforestation is performed based on the output data and online satellite imagery.

Finally, a complete alternative for the simulation of seawater intrusion on fully geospatial groundwater flow models of coastal aquifers based on open source software. The groundwater flow model was constructed on the Tupac Cloud platform with two stress periods and a total simulation time of 40 years. The Tupac project is downloaded and run locally on Anaconda where the Buy package for variable density flow together with the transport model are implemented on a Flopy script. A graphical representation of the grid, boundary conditions and results from the flow and transport models are developed as well on the Jupyter notebook.