# H06-E pick justification (one page)

**Date:** 2026-09-13T23:18:04Z UTC · **Hub job:** `77356229-193e-4693-a90c-cae8dcb0bc4a`  
**Ops:** H04-E MECHANICAL (STEO copies weekly crude stocks); H05 same class (NW2→NGWGPUS). James wants results that click — DIFFERENT cheap exploratory less likely mechanical inventory-copying.

## Chosen child
**H06-E** — NOAA Contiguous-US monthly **HDD+CDD** (Climate at a Glance) → STEO Table 7a retail electricity sales (`ELTCTWH` / pre-break `EXTCPUS`) revision across adjacent STEO issues.

## Why runnable today
| Leg | Source | Status |
|-----|--------|--------|
| Physical | NOAA NCEI CAG national HDD+CDD CSV | HTTP 200; HDD sha `424612de7c254ca3…`; CDD sha `3bd0c65be3b98604…` (hist→vintage unverified) |
| Outlook | STEO archives `*_base.xlsx` | **141/141** jan2015–sep2026 already on disk (shared H04/H05) |

## Why not other lanes this cycle
- **H04/H05 inventory→STEO:** MECHANICAL / likely mechanical — do not redo stocks.
- **H01 corn / FT900 / ESR:** blocked/killed / archive digs — James: do not dig.
- **EIA-930→STEO:** still first-print UNKNOWN; NOAA monthly DD is fetchable today without 930 archaeology.
- **Baker Hughes / DPR:** DPR xlsx fetchable but closer to STEO Table 10a mechanical risk; weather→sales is cleaner non-inventory click test.

## Test shape (H01-E / H04-E discipline)
Frozen baseline vs +one feature; RMSE; ≥5% bar; EXPLORATORY / CAPITAL PROHIBITED / CLAIM_STRENGTH LOW / MODEL_CHANGE=NO.

## Mechanism (hypothesis only)
Realized degree-day intensity between STEO issues → next official electricity retail-sales path may revise. Incremental predictiveness only — not capital.

## Mechanical risks (upfront)
1. STEO embeds weather (ZWHDPUS/ZWCDPUS) — sales revision may track STEO weather-table update.
2. Near-history sales cells may absorb Electric Power Monthly actuals.
3. NOAA timing SCHEDULE_PROXY; CAG hist can revise.
4. EXTCPUS→ELTCTWH code break ~Aug2019 (alias locked).