Zoom session to prepare all participants for the in-person workshop to make it as efficient as possible so that everybody is ready to go at the beginning of the workshop (February). Who has Windows/Ubuntu/Mac.
Welcome.
Presentation of participants - expectations.
Discussion on plan for the workshop - expectations. Organise into groups of 2 or 3.
Check that all necessary installations are in place (netCDF files with October-September total precipitation/wet-day frequency and wet-day mean precipitation; seasonal temperature files. Region: 10°E-60°E/45°S-5°N; ERA5 and a selection of GCMs in a shared Google folder). Check daily data with provided R-shiny app (same as https://ocdp.met.no/).
30 min Lecture: Introduction to downscaling: why and what is it? Focus on empirical-statistical downscaling (ESD) and weather generators (link).
15 min Lecture: Introduction to R-markdown scripts. Start with a template (link).
Start working with the data in groups. Calibrate the ESD method on ERA5 reanalysis and check the results.
Individual reflection/assessment of the day's progress. Describe the results - do they make any sense?
Discussion on the assignments.
30 min Lecture: How to make regional climate projections: calibration and evaluation (cross-validation) (link).
Work in different groups (new partner). Downscale one GCM simulation, evaluate the methods, and present the results.
Work in different groups. Downscale an ensemble of GCM simulations and present the results.
Answer the question: How well did the predictor capture the predictand? Were all stations equally well represented by the model?
Discuss the previous assignment.
30 min Lecture: Weather generators (WGs) (link)
Work in different groups. Calibrate a weather generator on a selected station.
Work in different groups. Apply the weather generator to downscaled rainfall.
The way forward: write a science paper - describe the method - collect results - collect literature. We use a Google document and co-write.
Decide next teleconference.