Speaker:       Dr. Lieke Melsen

Location:        Online

YouTube Live:     youtube.com/dicuchilefcfm

Date:           September 29, 2022

Time:           14:30 - 16:00 (CL time)

Abstract

Even if you give different people the same recipe and ingredients, the final dish will still taste
differently. The same applies to the use of computer models: different modelers will make
different choices, leading to different model results. In this study, I interviewed 14 modelers
to study their modeling choices. In total, 83 different reasons to make a certain decision were
identified. The most frequently mentioned reason to do something in a particular way, was
the experience from colleagues. Most reasons were context dependent, they were for
instance related to time constraints, available resources, colleagues, and personal
preferences. This makes model results time and place dependent. It is important to be aware
of this when estimating how reliable model results are.

Biography

Lieke Melsen works as assistant professor in the field of computational hydrology at Wageningen University, the Netherlands. Her research focuses on the selection and use of models, considering modelling as a socio-technical activity, and the consequences for model uncertainty. For this she combines technical approaches - model ensembles, sensitivity analysis - with quantitative research methods - interviews, ethnographic research.