Fix CO2 limits after the merge of power and sector models

[danielelerede-oet]

Checklist

• I am using the current main branch or the latest release. Please indicate.
• I am running on an up-to-date pypsa-earth environment. Update via conda env update -f envs/environment.yaml.

Describe the Bug

Using the current, default CO2 limit field from the config file may be dangerous as that value is tailored on the electricity sector, and anyway quite low for large/largely emitting countries and users may not notice/know that the value has to be changed (especially if they're running the sector model).

Here's the case of Nigeria when using the default CO2 limit value:

electricity generation (TWh)	Power	Sector
CCGT	18.44	0.13
solar	15.96	36.81
hydro	5.22	5.23
coal	2.05	0.00
onwind	0.00	0.00
biomass	–	183.52
solar rooftop	–	3.09

My proposal to overcome this issue is:

1. have a collection of CO2 limit values for e.g., 2020 for all the countries in the world for both the electricity generation / total energy-related emissions, so that the countries field is already sufficient to all the correct values
2. Implement a differentiation between electricity-related CO2 limit and total emissions from energy uses, reducing the users' effort to select the correct value of CO2limit

FYI @ekatef @davide-f

[ekatef]

@danielelerede-oet many thanks for opening the issue! A great point to address. Completely agree that the current design is not particularly user-friendly in the emissions part.

If I get your idea properly for point 1, you suggest an automated way to extract the emission level for each specific country from an emission database. To me, it sounds like a very convenient approach and I wonder if it may be possible to build it on top the infrastructure we still have in the model.

Currently, we have an option to calculate the emission limit automatically with the following options:

pypsa-earth/config.default.yaml

Lines 161 to 162 in 640f38a

	automatic_emission: false
	automatic_emission_base_year: 1990 # 1990 is taken as default. Any year from 1970 to 2018 can be selected.

Using automatic_emission: true triggers automated extraction of power sector emissions in prepare_network. You can check the details of the implementation in #748. EDGAR database is used as a data source. Currently, we extract only emission values which relate to Public electricity and heat production but the database itself is designed as a global emissions inventory and can be probably handy to fix the emissions treatment also for the global models.

What do you think?