Add 80m wind, temperature, and pressure to NOAA GEFS

[mrshll]

What

Adds four near-surface data variables at the 80 metre level to the shared NOAA GEFS data-variable config (src/reformatters/noaa/gefs/common_gefs_template_config.py):

• wind_u_80m, wind_v_80m (80 m U/V wind component)
• temperature_80m (80 m temperature)
• pressure_80m (80 m pressure)

All four come from the GEFS 0.5° "b" file (pgrb2b.0p50, gefs_file_type="b") — the same file type as the existing wind_u_100m/wind_v_100m.

Affected datasets (template metadata regenerated):

• noaa-gefs-forecast-35-day — gains the 80m level
• noaa-gefs-analysis — gains the 80m level
• noaa-gefs-forecast-10-day-spatial-dev — unaffected; it filters to _S_FILE_TYPES (0.25° s-files) and so excludes gefs_file_type="b" variables by design (just like the existing 100m wind). Its template is unchanged.

Why

Customer wind-energy use case: 80 m is a common wind-turbine hub height, so U/V wind plus temperature and pressure at that level support hub-height wind and air-density calculations.

Source verification

Source availability was verified against a live GEFS GRIB index: UGRD, VGRD, TMP, and PRES at 80 m above ground are all present in the 0.5° b file. NOAA byte-range matching keys off grib_element + grib_index_level (+ lead-time string), which the new configs set to match the live index. The 80m read/regrid path is the same b-file path already exercised in production by the existing 100m wind.

Metadata mirrors existing equivalents: wind mirrors wind_u_100m/wind_v_100m, temperature mirrors temperature_2m (K→°C conversion happens automatically on read), pressure mirrors pressure_surface but uses the generic-height CF standard_name="air_pressure" (not surface_air_pressure). keep_mantissa_bits: wind 6, temperature 7, pressure 10 (matching the dataset's pressure_surface).

CF-compliance exemptions

tests/common/datasets_cf_compliance_test.py validates long_name/short_name against the ECMWF parameter database. ECMWF has 80m U/V wind (80u/80v, already exempt for HRRR) but no 80m temperature or 80m pressure parameter, so 80t/80sp and their long names were added to the existing ECMWF_SHORTNAME_EXEMPT / ECMWF_LONGNAME_EXEMPT sets — the test's designed mechanism for quantities without an ECMWF entry. Without this the compliance test fails.

Verification

Verified on the canonical Python 3.14.6 toolchain: update-template for all three GEFS datasets produces zero diff vs. the checked-in metadata (spatial-dev unchanged, confirming the b-file filter); ruff / ty check / pytest (259 passed) all pass; and a decode of real GEFS GRIB produced physically sensible values. See the verification comment below.

Scope

Template metadata only. Post-merge work — history backfill, validation (incl. a K→°C sanity check on the b-file temperature), and catalog docs — is tracked in #752.

Merge-order note: #748 and #749 both add the 80t/80sp CF exemption and will conflict; whichever merges second must drop the duplicate lines.

🤖 Generated with Claude Code

https://claude.ai/code/session_015ZY5VagHhMYq9D6WXVXxvq

[mrshll]

Verification on canonical Python 3.14 toolchain

Re-verified on real 3.14.6 (uv sync, template regen, ruff/ty/pytest, and decode of real source GRIB).

• Template diff: regenerated noaa-gefs-forecast-35-day, noaa-gefs-analysis, and noaa-gefs-forecast-10-day-spatial-dev → zero diff vs. checked-in metadata. The spatial-dev dataset shows no change, confirming the new b-file (gefs_file_type="b") vars are correctly filtered out by _S_FILE_TYPES.

• ruff format / ruff check / ty check: all clean.

• common_template_config_subclasses_test + datasets_cf_compliance_test: 259 passed.

• Decode (GEFS forecast-35-day, member 0, first lead times) — all physically sensible, no NaN:

  variable	units	min	max
  temperature_80m	degree_Celsius	-54.25	40.5
  pressure_80m	Pa	49792	103104
  wind_u_80m	m s-1	-31.0	35.0
  wind_v_80m	m s-1	-33.0	35.5

No corrections needed — the branch is correct as-is.

⚠️ Merge-order dedupe with #748 — exemption kept here

This PR and #748 both add 80t / 80sp (+ long names) to the cf-compliance exempt sets. I tested removing them from this branch: datasets_cf_compliance_test then fails for noaa-gefs-analysis and noaa-gefs-forecast-35-day (the 80m temperature/pressure vars have no ECMWF param-DB entry). So the exemption must stay on this branch and cannot be pre-emptively dropped here.

Consequence: whichever of #748 / #749 merges second must drop the duplicate 80t / 80sp / 80 metre temperature / 80 metre pressure lines the first merge already added, to avoid a duplicate-key conflict.

Automated verification; PR left as draft for human review.