Update reference grids generation

[cjacobscrisioni]

We want to create updated 2021 population grid maps (1km) that are:

• Consistent in absolute terms with Ardeco adjusted population grids.
• Consistent in relative terms with municipal population counts by age and sex.

I suggest we start with using the census grids by sex and age groups (https://ec.europa.eu/eurostat/web/gisco/geodata/population-distribution/population-grids , version january 2025) to guide the initial disaggregation of the grids.

The Ardeco adjusted population grids should be available in the Networkmodel_EU sourcedata.

The municipal populations by age and sex are downloadable here: https://cjacobscrisioni.stack.storage/s/zdI8fkJTf0uffLJR as both csv and fss files. There is also a municipality geopackage that is joinable with the data throught the GISCO_ID field.

Caveat: my main worry is that some municipalities in the EU are quite small with capricious borders. We might want to move to a finer resolution in geographies on the border of municipalities.

Note that the current IGOR configuration assumes 100m population distributions

[cjacobscrisioni]

The municipal populations were used to assess net migration per age and sex. These describe cohort sizes in e.g. 2011 and 2021. Numbers on the oldest age classes are not available for the start year (no cohort left in 2021) and idem numbers on the youngest age classes are not available for the end year (youngest moved to the third cohort by then). On second thought we may have better data to describe municipal population, will query.

[cjacobscrisioni]

Better data should be available through https://territorial.ec.europa.eu/ardeco/explorer?lng=en but the site seems down right now. Will check again on Monday.

[cjacobscrisioni]

Received better municipal data in fst format: https://cjacobscrisioni.stack.storage/s/oT2bnwqdTcDRU3hp. Joinable through the GISCO_ID field.

The fst format is readable through R with the fst library. Alternatively I can export them as e.g. csv files or similar. @jipclaassens please let me know which format you prefer :)

[jipclaassens]

The fst format is readable through R with the fst library. Alternatively I can export them as e.g. csv files or similar. @jipclaassens please let me know which format you prefer :)

If it is not much work, any other useable format would be preferable. As i otherwise would need to install R, figure out how it works, etc.

.csv or even more efficient: .parquet

[cjacobscrisioni]

The fst format is readable through R with the fst library. Alternatively I can export them as e.g. csv files or similar. @jipclaassens please let me know which format you prefer :)

If it is not much work, any other useable format would be preferable. As i otherwise would need to install R, figure out how it works, etc.

.csv or even more efficient: .parquet

@jipclaassens , csv files here: https://cjacobscrisioni.stack.storage/s/Gv27BjCyf8g79iD3 (I will have to look into exporting to parquet another time)