# The pack for incident.datascience@web-summit.dinis > Kept by incident.librarian@web-summit.dinis, 2026-11-11, 12:18 UTC: written by `tools/library-pack.mjs` on every push from the files it points at. Read [start.md](start.md) first, then this, then open only what you need. A pointer is `path` (bytes, the first 10 hex digits of its sha256); any path opens at `view/?path=`. [The pack](README.md). ## Your job **data science** (Dinis, 11 Nov ~11:45, BRIEF.md): the gap map against Watchfloor's decision brief (research/watchfloor/README.md); new open data sets, each read at its own source with its licence: people exposed by district and hour (Kontur/INE on IPMA's warning areas), the ECMWF ensemble, tides, low land, hospitals, power; how to communicate it (layers from simple to deep). (from `agents/JOBS.md` (16,087 B, cb88df47c9)) **You may change:** `datascience/**`, `incidents/*/data/exposure/**`, `tools/ds-*.mjs`, `agents/datascience/**`, `accounts/ledger/datascience.jsonl`, `transition/interviews/datascience.md`, and your own entry in `data/keys.json`. (from `agents/policy.json` (13,565 B, ec9b7b8d15)) **Your draft behaviour policy** (not agreed by Dinis): `agents/datascience/policy.md` (8,553 B, a8284ca4c1) ## Where the work stands - **State:** first round done: gap map; exposure data sets (population, people by hour, tides model, OSM hospitals, Lisbon metro and 2 km squares with low land, ECMWF ensemble crossings); layers.md (its own date: 2026-11-11T11:59:00Z) - **Next:** refresh people-by-hour after each IPMA change; ensemble after each ECMWF run - **Next:** power (E-REDES) and ANEPC counts: find an open feed - **Next:** precedents table (IPMA storm reports) - **Ask:** Dinis: Instituto Hidrográfico's official tide tables are sold through loja.hidrografico.pt (account needed); tides.json is a model until then - **Ask:** Dinis: overpass-api.de, tabulae.ine.pt, download.geofabrik.de have no route from this session (worked around with HOT's OSM export and INE's indicator API) - **Ask:** done: raw inputs in vault yiqem57n, key sealed to the conductor via the lane (DEM tiles not uploaded: public, sha256 listed) - The whole status, with everything done: `agents/datascience/status.json` (1,239 B, 2c49d6687e) **Open first, before any other file: the hand-over written by the session before yours:** `datascience/HANDOVER.md` (1,819 B, 504e1ff7aa) ## Your files 16 files, 269,832 bytes in all (about 112,430 tokens to read every one): `incidents/2026-11-10-storm-charles/` 8, `datascience/` 3, `tools/` 3, `datascience/tools/` 2. The ones to open first: - `tools/ds-ensemble.mjs` (5,797 B, 69bfc1811f): incident.datascience@web-summit.dinis: where the storm's centre crosses Lisbon's meridian in each of the 51 … - `tools/ds-exposure.mjs` (5,106 B, 973c640c85): incident.datascience@web-summit.dinis: residents under each IPMA warning level, by warning area and hour. … - `tools/ds-tides.mjs` (4,554 B, 862532306d): incident.datascience@web-summit.dinis: high and low water at Cascais (Tagus mouth), Lisbon and Funchal, from … - `incidents/2026-11-10-storm-charles/data/exposure/README.md` (4,032 B, 0ad318d0e0): Not a warning service: IPMA issues the warnings and ANEPC the instructions. These files say where people live … - `incidents/2026-11-10-storm-charles/data/exposure/ensemble.json` (17,375 B, c8f7ca270a): Where the storm's centre crosses 9.5 W (off Lisbon) in each of the 51 ECMWF ensemble forecasts, grouped north … - `incidents/2026-11-10-storm-charles/data/exposure/hospitals.json` (100,691 B, d741ed86c0): Hospitals mapped in OpenStreetMap in Portugal (amenity=hospital or healthcare=hospital), each placed in its … - `incidents/2026-11-10-storm-charles/data/exposure/lisbon-grid.json` (12,839 B, 0d834d8b40): Lisbon in 49 squares of 2 km (7 x 7, rows A to G from north, columns 1 to 7 from west, centred on 38.745 N, … - `incidents/2026-11-10-storm-charles/data/exposure/lisbon-municipalities.json` (6,788 B, ac3c7bd564): The 18 municipalities of the Lisbon metropolitan area: residents, aged 65+ and 0-14 (Census 2021), land 0.5 … - `incidents/2026-11-10-storm-charles/data/exposure/people-by-hour.json` (13,004 B, eaf420face): Residents (INE Census 2021) living in each IPMA warning area, by hour, under the highest level in force there … - `incidents/2026-11-10-storm-charles/data/exposure/population.json` (39,109 B, fb437a9bbc): Residents counted by the 2021 Census, by IPMA warning area (district; Madeira by zone) and by municipality: … Every one, with who last changed it: [the catalogue](../index.json), filtered by your paths. ## Your neighbours What your files say they read, and who reads yours: taken from the links and sources the files themselves declare. A contract here is an address and a shape that does not depend on anyone's session. **You read:** - **incident.broadcast@** (1 file of theirs): `broadcast/data/field.json` (58,504 B, 870e77ea75): A model's wind (10 m) and sea-level pressure over the north-east … **They read yours:** - **incident.conductor@** (2 files of yours): `incidents/2026-11-10-storm-charles/data/exposure/README.md` (4,032 B, 0ad318d0e0): Not a warning service: IPMA issues the warnings and ANEPC the …; `incidents/2026-11-10-storm-charles/data/exposure/ensemble.json` (17,375 B, c8f7ca270a): Where the storm's centre crosses 9.5 W (off Lisbon) in each of the 51 … ## Open cards on the board None that name you as owner or wait on you. ## What this pack cost you 5,484 bytes (about 2,285 tokens, estimated at 2.4 bytes a token). It points at 16 files of 309,062 bytes; your own files are 269,832 bytes.