Mexico Reservoir Transparency

REGRESSION (verified Jun 2026). The structured daily publication this dimension previously rewarded has STOPPED. CONAGUA's SINA 'Monitoreo de Presas' feed has not produced a new reading since 23 April 2025: direct retrieval of the portal's own report API (PresasPG/presas/reporte/<date>) returns the full 210-dam array only for 2025-04-23, while every later date — including the day before retrieval — returns 0 records, and the app's 'latest date' endpoint itself reports 2025-04-23. The lapse is documented in Mexican media (Letra Fría, Diario de Chiapas, mid-2025). What remains publicly available is a 14-month-old frozen snapshot plus the annual Estadísticas del Agua, not the daily structured feed the prior score assumed.

The channels that earned the prior 52 are non-functional. The REST report endpoint (sinav30.conagua.gob.mx:8080/PresasPG/presas/reporte/) serves only frozen 2025-04-23 data. The SIH FTP server (ftp://presas:presas@sih.conagua.gob.mx/) that provided per-station CSV downloads is dead: sih.conagua.gob.mx now resolves to an Imperva Incapsula WAF (HTTPS returns X-Iinfo / incap_ses cookies, IP 45.60.76.106) that does not route FTP — port 21 accepts the TCP connection but returns no FTP banner (verified plain-FTP and FTPS, Jun 2026). The public GeoServer (geosinav30) still answers but exposes dam metadata/location layers, not a current storage feed. Net: no working live channel for current dam storage.

Capacity-weighted on CURRENTLY-OBTAINABLE data — the same standard applied to India. With the SINA daily feed frozen at 23 April 2025 and the SIH FTP archive dead, essentially none of Mexico's ~165,000 hm³ of national reservoir capacity is publicly trackable at a current reading. A stale 14-month-old snapshot of the 210 principal dams is technically retrievable, but a frozen value is not 'trackable storage'. The previous score of 88 reflected a daily feed that had, in fact, already been frozen for ~13 months at the time of the May-2026 evaluation. (Untested possible live alternatives — CFE for major hydro, state water commissions — could restore some coverage if integrated; none is confirmed.)

The deep archive that justified the prior 95 — the SIH FTP per-station series back to 1991 — is no longer reachable (the FTP host is behind a WAF that does not route FTP). What survives is the web report API's history from 2007 up to the 23 April 2025 freeze (~18 years) for the principal-dam set, plus bulk historical datasets on Datos Abiertos. Substantial history therefore remains accessible, but it is truncated (ends ~14 months ago) and the deepest channel is gone, so the accessible median depth drops from the prior 'median 30+ years' band.

The public feed has not updated in ~14 months — last reading 23 April 2025, verified by direct retrieval in June 2026 (every post-April-2025 date returns 0 records). The portal still advertises daily updates but delivers none. This is the dimension most directly broken by the publication lapse: a system designed for daily cadence that has gone dark.

Largely unchanged in principle: capacity figures (NAMO/NAME storage and crest/operating elevations) for the 210 principal dams and the national dam-registry methodology were documented in the Anuario Estadístico del Agua and embedded in the SINA/GeoServer attributes. But the per-dam current figures that anchored those definitions now exist only in the frozen 2025-04-23 snapshot, and there is still no single public document explaining gauge-to-volume conversion curves, bathymetric survey dates or sensor calibration — so practical methodological transparency slips slightly from the prior 60.

Unchanged. All CONAGUA portals, reports and data systems remain exclusively in Spanish — the SINA portal, the (now-dead) SIH FTP file names, the GeoServer layer names, the Datos Abiertos metadata and the weekly PDF reports are Spanish-only with no English translation layer or data dictionary.