airmosaic-ai

Sector Naming Conventions in GCAM-China

Internal names, mappings, and conventions that agents must know before writing queries or modifying configurations.

Sector Names vs Display Names

GCAM-China uses internal sector names that may differ from human-readable names:

Internal Name	Display / Common Name	Sector File
`ammonia`	N fertilizer / nitrogen fertilizer	`Fert_CHINA.xml`
`cement`	cement	`cement_CHINA.xml`
`iron and steel`	iron and steel	`detailed_industry_CHINA.xml`
`chemical`	chemical energy use	`detailed_industry_CHINA.xml`
`coke`	coke	`detailed_industry_CHINA.xml`
`other industry`	other industry	`detailed_industry_CHINA.xml`
`aluminum`	aluminum	`detailed_industry_CHINA.xml`
`paper`	paper	`detailed_industry_CHINA.xml`
`food processing`	food processing	`detailed_industry_CHINA.xml`

Always use the internal name in XQuery. The display name may appear in documentation or output labels but is not the queryable identifier.

The `ammonia` / `N fertilizer` Mapping

The R constant confirms the internal name:

# From input/gcamdata/R/constants.R
gcamchina.FERT_NAME <- "ammonia"

XML files generated by GCAM-China R scripts use ammonia, but some externally sourced or older CCS files may use N fertilizer. When in doubt, query the database:

distinct-values(collection()/scenario/world/*/*[@type="sector"]/@name)

Province Codes

GCAM-China uses two-letter abbreviation codes for Chinese provinces:

AH, BJ, CQ, FJ, GS, GD, GX, GZ, HI, HE, HL, HA,
HB, HN, NM, JS, JX, JL, LN, NX, QH, SN, SD, SH,
SX, SC, TJ, XZ, XJ, YN, ZJ

The China node is the national aggregate, not a province. It is treated as a region but serves a different role:

China aggregate: national-level sector definitions, pass-through sectors for energy distribution, national account
Province nodes: province-level production sectors with their own technologies, costs, and outputs

Province Code Mapping File

Located at database_R/Prov_code_china.csv, with columns:

Province_code: numeric code
EN_name: English province name (e.g., Anhui)
GCAM_name: two-letter abbreviation used in GCAM (e.g., AH)

Technology Naming Patterns

Technologies follow these patterns:

Pattern	Example	Sector
Base name only	`cement`	cement
Base + `CCS`	`cement CCS`	cement with CCS
Process variant	`BLASTFUR`, `EAF with DRI`	iron and steel
Feedstock variant	`coal`, `gas`, `hydrogen`	ammonia

Technology names include spaces and are case-sensitive in XQuery.

Node Type Names

GCAM XML nodes use specific @type values:

@type	Element	Description
`region`	region	Top-level region node
`sector`	supplysector, pass-through-sector	Sector node
`subsector`	subsector	Subsector node
`technology`	technology, stub-technology	Technology node
`output`	output-primary, output	Output node
`input`	input	Input node
`resource`	resource, renewresource, unlimited-resource	Resource node

Use @type in XPath for reliable node selection:

$r/*[@type="sector" and @name="cement"]

This is more robust than matching on element name because GCAM-China uses both supplysector and pass-through-sector for sectors.

Common Query Pitfalls

Using display names instead of internal names: Query ammonia, not N fertilizer
Querying China node for province data: The China node has national aggregate values, not province-level data
Assuming all provinces have all sectors: Not all sectors exist in every province
Case sensitivity: "Cement" != "cement" in XQuery string comparisons

This site is open source. Improve this page.