Clean up, update, and expand geo mapping files

[jtlangevin]

Updated all the mapping data here and reflected them in the Scout geo mapping files and baseline data, on top of @trynthink's fixes for the commercial data.

[jtlangevin]

@trynthink a note that I see some small negative values for health care building type in a few regions, and would appreciate if you can verify those negatives are present in the underlying EIA data.

e.g., for new england -> health care -> electricity -> MELs -> other:

"other": {
            "energy": {
              "2016": -1151449.0,
              "2017": -1288993.0,
              "2018": -1464411.0,
              "2019": -1444773.0,
              "2020": -113440.0,
              "2021": -120260.0,
              "2022": -124197.0,
              "2023": -156029.0,
              "2024": -155408.0,
              "2025": -152845.0,
              "2026": -150379.0,
              "2027": -149298.0,
              "2028": -148734.0,
              "2029": -148257.0,
              "2030": -147598.0,
              "2031": -147358.0,
              "2032": -147359.0,
              "2033": -148382.0,
              "2034": -158456.0,
              "2035": -158289.0,
              "2036": -166320.0,
              "2037": -167073.0,
              "2038": -172429.0,
              "2039": -173035.0,
              "2040": -173251.0,
              "2041": -173545.0,
              "2042": -173434.0,
              "2043": -173368.0,
              "2044": -172478.0,
              "2045": -171715.0,
              "2046": -171437.0,
              "2047": -171040.0,
              "2048": -171027.0,
              "2049": -169733.0,
              "2050": -168722.0
            },
            "stock": "NA"
          }

[trynthink]

@trynthink a note that I see some small negative values for health care building type in a few regions, and would appreciate if you can verify those negatives are present in the underlying EIA data.

e.g., for new england -> health care -> electricity -> MELs -> other:

"other": {
            "energy": {
              "2016": -1151449.0,
              "2017": -1288993.0,
              "2018": -1464411.0,
              "2019": -1444773.0,
              "2020": -113440.0,
              "2021": -120260.0,
              "2022": -124197.0,
              "2023": -156029.0,
              "2024": -155408.0,
              "2025": -152845.0,
              "2026": -150379.0,
              "2027": -149298.0,
              "2028": -148734.0,
              "2029": -148257.0,
              "2030": -147598.0,
              "2031": -147358.0,
              "2032": -147359.0,
              "2033": -148382.0,
              "2034": -158456.0,
              "2035": -158289.0,
              "2036": -166320.0,
              "2037": -167073.0,
              "2038": -172429.0,
              "2039": -173035.0,
              "2040": -173251.0,
              "2041": -173545.0,
              "2042": -173434.0,
              "2043": -173368.0,
              "2044": -172478.0,
              "2045": -171715.0,
              "2046": -171437.0,
              "2047": -171040.0,
              "2048": -171027.0,
              "2049": -169733.0,
              "2050": -168722.0
            },
            "stock": "NA"
          }

I did some spot checking and these negative values are consistent with the difference in the values reported by EIA in the commercial microdata files for total electric other energy use and the sum of all MELs energy use. I can't explain why there would be negative values like this, but it is consistent with the AEO data.

[trynthink]

@jtlangevin With this update to the geography mapping spreadsheet, there are eight geography translation files that are not included but continue to be used by final_mseg_converter.py.

1. Com_Cdiv_Czone_ColSums.txt
2. Com_Cdiv_Czone_RowSums.txt
3. Com_Cdiv_EMM_Elec_EU_RowSums.csv
4. Com_Cdiv_State_Elec_EU_RowSums.csv
5. Res_Cdiv_Czone_ColSums.txt
6. Res_Cdiv_Czone_RowSums.txt
7. Res_Cdiv_EMM_Elec_EU_RowSums.csv
8. Res_Cdiv_State_Elec_EU_RowSums.csv

How are these files updated? Do they need to be added as sheets in the spreadsheet?

[jtlangevin]

@trynthink that's right.

The "EU" ones are produced via post-processing EULP data (@handichan can comment more on the specific method used here.) So that's coming from a separate process and dataset.

Com_Cdiv_EMM_Elec_EU_RowSums.csv Com_Cdiv_State_Elec_EU_RowSums.csv Res_Cdiv_EMM_Elec_EU_RowSums.csv Res_Cdiv_State_Elec_EU_RowSums.csv

The "_CdivCzone" ones must have been produced from older RECS/CBECS data that included the AIA czone distinction, because the factors differ from residential vs. commercial and therefore can't just be derived from county-level population estimates (which span both building types). I wasn't able to tie them back to data we have in this spreadsheet, at least.