GeoTwinn automated calibration of groundwater flow models training for Work package 2.1

GeoTwinn

GeoTwinn automated

calibration of

groundwater

flow

models training

for

Work package 2.1

VENUE: Croatian geological Survey (HGI-CGS), Milana Sachsa 2, Zagreb, CRO
Attendees: Željka Brkić (HGI-CGS), Maja Briški (HGI-CGS) and Ozren Larva (HGI-CGS)
GEUS experts: Simon Stiesen (GEUS), Julian Koch (GEUS)
Duration: August, 26 – August, 28 2019

Within the framework of GeoTwinn H2020 project, another training for scientists from Croatian Geological Survey (HGI-CGS) was organised. The course on automated calibration of groundwater flow models was held by two experts from Geological Survey of Denmark and Greenland (GEUS): Simon Stiesen and Julian Koch.

During the first day of the training HGI-CGS staff presented the results of activities conducted so far in terms of groundwater flow modelling of the Zagreb pilot area. It was followed by in-depth analysis of the model grid structure relative to geological model, groundwater flow boundaries and results, including the analysis of residuals. Also, potential causes of numerical model convergence issues were thoroughly examined.

The second day was dedicated to application of different strategies of automated calibration using PEST in order to improve the performance of the model and reduce residuals. GEUS staff demonstrated the application of the methods and procedures of setting up and running PEST outside the MODFLOW modelling environment (GMS) in order to improve efficiency of calibration process. Besides, GEUS experts showed how combination of running parallel PEST and application of custom-made Python scripts can improve and speed up calibration process.

During the last day HGI-CGS and GEUS staff continued the work on automatic calibration of the model. It was also demonstrated how application of Python scripts can lead to better visualization of the modelling output. In the afternoon, the results of Copernicus EDgE project were discussed in terms of their applicability for the purpose of propagation of the climate change consequences through calibrated groundwater flow model of the Zagreb pilot area. It was followed by discussion on strategies for the model improvement in the period until the next visit of HGI-CGS staff to GEUS in October, 2019.

Hereby we would also like to express our gratitude to the lecturers, with whom it was a pleasure working.