Meteorological Intelligence (METOCINT): Intelligence Discipline Guide

Weather is the one adversary that files its own forecast. Meteorological intelligence turns that forecast into decisions about people, assets and operations.

What Meteorological Intelligence is as a discipline

Meteorological and oceanographic intelligence is the collection and analysis of atmospheric, marine and hydrological conditions in support of operational decisions. It ingests numerical weather prediction output, surface and upper-air observations, radar and satellite products, lightning detection, buoy and tide-gauge data, sea state and current models, and official warnings. The practice is less about predicting weather, which national services already do well, than about translating meteorological data into consequence: which sites lose power, which routes close, which shipments slip, and when the decision point actually arrives.

Sub-methods include deterministic forecast interpretation, ensemble and probabilistic analysis for risk framing, nowcasting from radar and lightning networks for short-fuse decisions, and climatological analysis for planning and site selection. Maturity progresses from consuming a public forecast, to threshold-based alerting against documented operational tolerances, to full decision-support products with defined lead times and trigger points agreed with operations before the season begins rather than during an event.

Why it matters

Meteorological intelligence answers when you have to decide and what happens if you do not. A forecast alone changes nothing; a forecast tied to a specific operational threshold and a defined lead time changes behaviour. It supports evacuation and shutdown timing, maritime and aviation routing, physical security planning for outdoor sites and events, supply chain disruption forecasting, and the reconstruction of conditions during an incident for insurance and legal purposes.

What analysts actually look for

These are the concrete, observable signals that carry weight in this area of work:

Model forecasts of wind speed and gusts, precipitation rate and accumulation, temperature extremes and visibility at specific asset locations
Ensemble spread and probabilistic exceedance of operational thresholds, expressing forecast confidence rather than a single deterministic value
Radar reflectivity, storm-cell tracking and lightning detection supporting nowcasts with lead times of minutes to a few hours
Tropical cyclone track and intensity forecasts with uncertainty cones, landfall timing and wind-field radii relative to facilities
Sea state including significant wave height, period and direction, swell, currents, tides and storm surge levels at ports
River gauge levels, soil moisture and hydrological forecasts driving flood onset timing for inland facilities and transport routes
Official warnings, watches and advisories issued by the responsible national authority, with their published trigger criteria
Verified historical observations used to reconstruct exact conditions at a place and time during an incident or claim

Where the data comes from

Authoritative and openly available collection points. Always confirm licensing and terms before operational or commercial use:

NOAA and the National Weather Service — Free authoritative forecasts, warnings, model output and archived observations covering the United States
ECMWF open data — Global deterministic and ensemble forecast output, widely regarded as the leading medium-range guidance
NOAA NDBC buoys — Real-time and archived marine observations of wave height, wind, pressure and sea surface temperature
National Hurricane Center — Official tropical cyclone track, intensity and wind-probability products with defined uncertainty cones
Copernicus Marine Service — Ocean current, sea state and sea level analysis and forecast products for maritime operations
NOAA NCEI and Meteostat — Archived station observations for historical reconstruction, climatology and threshold calibration
METAR and TAF feeds — Standardised airport observations and terminal forecasts providing dense, frequent ground truth
National meteorological services — Met Office, DWD and JMA issue the authoritative regional warnings and regional model products

A working method

A repeatable sequence beats ad-hoc searching. This is a practical starting workflow:

Define operational thresholds — Agree with operations the exact conditions that trigger each action, per site and per activity, before any weather event occurs.
Set required lead times — Work backwards from how long each protective action takes, so alerts fire while the decision is still executable.
Ingest multiple models — Compare deterministic runs from independent centres and read the ensemble spread. Agreement and disagreement both carry information.
Localise the forecast — Adjust model output for terrain, coastal effect and site exposure, because grid-scale output rarely matches conditions at one facility.
Alert on thresholds — Report the probability of exceeding the agreed threshold within a stated time window, not a general description of the weather.
Nowcast through the event — Switch to radar, lightning and live observations as lead time collapses, and update the recommendation as conditions evolve.
Verify afterwards — Score forecasts against observed conditions and archive the record for insurance, claims and future threshold tuning.

How this connects across the intelligence taxonomy

Intelligence work does not respect neat boundaries. The mission domain you are working, the disciplines you practise, and the data points you pivot on are one connected system. These are the direct relationships for this entry — every link is also a tag, so you can follow any thread across the whole library.

Applied in these mission domains

Operates on these data points

Location / Coordinates — A geographic point, place, or region — the basis of GEOINT analysis.
Satellite Imagery — Overhead imagery of an area of interest, used for change detection and site analysis.
Event / Incident — A discrete real-world occurrence — protest, strike, breach, seizure — with time, place, and actors.
Facility / Site — A physical installation — plant, base, port, data centre — with a fixed footprint and function.
GPS Coordinates — Precise latitude/longitude coordinates identifying an exact point on Earth — the atomic unit of GEOINT analysi
Company / Organization — A legal entity — corporation, LLC, NGO, or business.

Related disciplines

Environmental Intelligence — Environmental Conditions, Damage, and Crime

Inside the platform: where Meteorological Intelligence lives

The Quantus platform is 204 pages behind a 147-item sidebar organised into six working groups: Command (24 items), Dashboards (15), Threat Theaters (14), Intelligence Domains (15), Investigate (34), and Administration (45). This entry is not a page in isolation — it is a thread running through several of them.

The modules that matter most here:

discipline.php?d=METOCINT — Discipline hub
source-catalog.php?disc=METOCINT — Source catalogue filtered to this discipline
search.php — Advanced search, filter and pivot
correlate.php — Correlation graph
cases.php — Case management

Each dashboard is local-first: it renders from the platform’s own database rather than depending on a live third-party call, so it still works when an upstream API is unreachable or rate-limited. Heavy aggregates are cached with a hard query time cap and degrade to the last good value instead of hanging the page.

Automation, playbooks and AI skills

Analysis that only happens when someone remembers to run it is not a capability. The platform ships a 30-step automation pipeline (cron.php) that collects, ingests, resolves, enriches, correlates and scores on a schedule — 25 seeders, 11 resolvers and 7 enrichment runners, all idempotent and cursor-based so a run can be interrupted and resumed without duplicating or losing work.

AI skills that apply

The 16 one-click operations in ai-skills.php are deterministic jobs, not free-text generation. The ones that matter here:

Score Country Risk
Sync Intel Domains
Resolve Everything
Summarise (Copilot)
Generate Report

Alerting closes the loop: rules in alerts.php fire on new indicators matching a saved query, so a first sighting in this area raises a notification rather than waiting to be noticed at the next review.

Feeds, data sources and the API

The collection layer runs a feed registry of free, machine-readable sources — bulk blocklists and trackers (Maltrail, IPsum, FireHOL, the full abuse.ch corpora, phishing databases, Emerging Threats, Spamhaus, DigitalSide, ThreatView), authoritative government feeds (CISA KEV, OFAC, UN and EU sanctions lists), and reference datasets (RIR allocations, ip-to-ASN and geolocation tables, MITRE ATT&CK, EPSS). collect.php pulls them server-side on a schedule; feeds.php and source-catalog.php show what is registered, what it covers and when it last ran.

Anything the platform holds is reachable programmatically. The REST API in api.php exposes 11 endpoints — status, stats, search, lookup, recent, export, bulk_check, top_threats, by_category, categories, check — and export.php streams 18 formats in bounded chunks, so a million-row export neither exhausts memory nor times out:

STIX 2.1, MISP, OpenIOC 1.1, CEF (ArcSight), LEEF 2.0 (QRadar), Zeek/Bro intel, Snort/Suricata rules, Palo Alto EDL, BIND RPZ, hosts blackhole, iptables, CSV, JSON, NDJSON/JSONL, XML.

That covers the CTI standards (STIX 2.1, MISP, OpenIOC), SIEM ingestion (CEF, LEEF, Zeek), detection engines (Snort/Suricata), and direct enforcement (Palo Alto EDL, BIND RPZ, hosts, iptables) — so intelligence developed here can be actioned in the tools you already run, without a manual reformatting step. A TAXII 2.1 server and a MISP/RSS feed are also served for pull-based sharing.

Use cases

Three ways this entry earns its keep in day-to-day work:

Triage under time pressure. An artifact or report lands and you need a defensible read in minutes, not days. Define operational thresholds is the first move; the platform pre-computes the enrichment so the analyst spends the time on judgement rather than lookups.
Building the picture. A single indicator is rarely the story. Ingest multiple models turns one artifact into a network — shared infrastructure, repeated selectors, the same operator behind different names — via the correlation graph and the cross-entity link engine.
Producing something actionable. Analysis that ends in a document nobody can use is wasted. Verify afterwards feeds the case file, the detection rule, the block list or the referral — with sourcing attached so the recipient can verify it.

Case management (cases.php), watchlists, saved searches and scheduled reports mean the work persists between sessions and survives an analyst leaving the team.

How each sector uses Meteorological Intelligence

The same entry is worked very differently depending on who you are, what authority you hold, and what you are ultimately producing. A military analyst is supporting a commander’s decision; a journalist is meeting a publication standard; an NGO caseworker is protecting a person. The underlying artifacts are shared — the constraints, outputs and thresholds are not.

🎖 Military and defence

Defence practitioners run meteorological and oceanographic intelligence as an operational support function with a formal place in planning. Products are decision-oriented: whether aviation can fly, whether a maritime insertion is within sea state limits, whether visibility and illumination favour an operation, whether a route will remain trafficable, and how conditions affect sensor and weapon performance. The discipline works from numerical weather prediction output, observations, and specialist products for illumination, refraction and acoustic propagation. It feeds mission planning, go and no-go decisions and force protection warnings for severe weather at fixed sites. Constraint: forecast uncertainty must be communicated as probability and threshold, because a commander needs the risk, not a single deterministic value.

🕵 National intelligence

National services use environmental conditions as an analytic input rather than an end in itself: assessing whether an observed activity was feasible on a given date, whether imagery gaps are explained by cloud, whether a claimed event is consistent with recorded conditions, and how weather constrains an adversary operating pattern. Reanalysis datasets provide an authoritative record of past conditions anywhere on Earth, which is powerful for testing claims. Fusion with imagery and signals reporting is routine. Handling is straightforward because the underlying data is open, but the requirement it serves and the location of interest are usually not, so tasking is protected while the meteorology is releasable.

👮 Law enforcement

Investigators use recorded weather and sea conditions as objective corroboration in incident reconstruction: road traffic collisions, maritime incidents, aviation accidents, fire investigation, search and rescue reviews and time-of-death estimation. The evidential requirement is to use the official archived observation or reanalysis for the specific location and time, obtained from the national meteorological authority with a certificate or documented provenance, rather than a consumer weather application. Distinguish observed conditions at a station from modelled conditions at the incident location, and state the distance and interpolation involved, because that is the point an expert witness will contest.

🔍 Private investigation and corporate security

Corporate practitioners use meteorological intelligence for business continuity, site risk, insurance claims and supply chain planning. Typical questions are whether a site is exposed to flooding, wind or heat thresholds that will interrupt operations, whether a claimed weather-related loss is consistent with recorded conditions, and how seasonal patterns should shape inventory and logistics decisions. Sources are entirely open and no special authority is needed. What matters is competence: interpreting probabilistic forecasts correctly, using the right product for the timescale, and not presenting a model output as an observation. Claims work in particular requires archived official records rather than reconstructed estimates.

📰 Journalism and OSINT media

Journalists use meteorological data to report on extreme events, attribution to climate change, and failures of warning and response. Verification means citing the official observation or forecast with its issuing authority, time and location, and distinguishing between a forecast, an observation and a reanalysis, which are routinely conflated. Single-event attribution to climate change is a specific scientific method and should be sourced to a rapid attribution study rather than asserted. Where reporting concerns a warning failure, obtain the actual warning text and timing rather than relying on recollection, and give the meteorological authority an opportunity to explain.

🌍 NGO, humanitarian and human rights

Humanitarian organisations use meteorological intelligence for anticipatory action, response planning and staff safety. The operational value is in the lead time: a tropical cyclone track forecast, a flood warning or a seasonal rainfall outlook triggers prepositioning, cash transfers and evacuation support days or weeks before impact, which is far cheaper and more effective than reacting. Practice requires understanding forecast probability and acting on defined triggers agreed in advance rather than on individual judgement in the moment. Duty of care to field staff means weather thresholds for movement and evacuation should be written, monitored and enforced rather than negotiated during an event.

🎓 University and research

Researchers use meteorological and oceanographic data for climate, hydrology, energy, agriculture and risk studies. Methodology must specify the dataset, its version, its spatial and temporal resolution and its known biases, since reanalyses and station records differ systematically and results are sensitive to the choice. Reproducibility requires archiving the exact extract and the processing chain, because datasets are reprocessed and versions superseded. Ethics review is rarely required for the physical data but becomes relevant where analysis is combined with population or health data on identifiable communities. Cite the dataset digital object identifier and version rather than the download interface.

Playbook: working Meteorological Intelligence end to end

A repeatable sequence, from the moment the requirement lands to the moment a product is delivered and the case is closed out. Each phase states what you are trying to establish, not merely what to click — the point is a defensible chain of reasoning, not a checklist.

Phase 1 — Define the decision and its threshold

Establish what decision the forecast supports and at what threshold conditions change it: the wind speed above which a crane stops, the sea state above which a transfer is unsafe, the visibility below which an approach is not flown, the rainfall above which a route floods. Without a threshold, a forecast is commentary. Capture the tolerance for a false alarm against a missed event, since that determines how the probability is used. A good output is a written decision statement with a numeric threshold and an owner. Stop when the threshold is specific enough to be monitored automatically.

Phase 2 — Establish the timescale and choose products

Match the product to the lead time. Nowcasting from radar and satellite covers the next few hours; high-resolution limited area models cover one to two days; global models cover several days to about a week with declining skill; ensemble and seasonal products cover longer horizons probabilistically and should never be read deterministically. Using a ten-day deterministic forecast as though it were a prediction is the standard failure. A good output is a product plan matched to each decision point. Stop when each decision has a product with demonstrated skill at that lead time.

Phase 3 — Characterise the location

Establish the local factors that models at coarse resolution cannot resolve: terrain channelling of wind, coastal and lake effects, urban heat, valley fog and cold pooling, exposure and aspect, and local drainage and flood history. Model output at a grid scale of several kilometres does not represent a specific site, and the difference is frequently the whole decision. Collect local station records to quantify the bias. A good output is a site characterisation with documented local corrections. Stop when you can state how the site typically differs from the nearest model grid point.

Phase 4 — Assemble observations and the climatological baseline

Gather current and historical observations for the location: surface stations, buoys and tide gauges, radar and satellite products, and lightning detection. Build the climatological baseline from a long record so that any forecast condition can be placed against what is normal for that location and season. Without a baseline, a forecast number carries no meaning for a decision maker. A good output is a baseline showing normal ranges and historical extremes for the relevant variables. Stop when the baseline is long enough to characterise the extremes that matter, not merely the average.

Phase 5 — Ingest deterministic and ensemble forecasts

Pull forecasts from more than one authoritative source and compare, since agreement between independent models is itself a confidence signal and disagreement is a warning. Prefer ensemble output wherever a decision has asymmetric consequences, because the spread carries the information about risk that a single run cannot. Record model run times and initialisation, since a forecast is only as current as its run. A good output is a forecast set with model, run time and lead time recorded for each. Stop when the ensemble spread has been examined rather than only the ensemble mean.

Phase 6 — Convert to probability against thresholds

Translate the forecast into the terms of the decision: the probability that the threshold is exceeded during the relevant window, rather than a single expected value. Use ensemble member counts or calibrated probabilistic products, and correct for known site bias. Present in the form the decision maker uses, whether that is a percentage, a category, or a simple exceed or not exceed with a confidence. A good output is a threshold exceedance probability with a stated window. Stop when the product answers the decision question directly rather than requiring the reader to interpret meteorology.

Phase 7 — Assess derived and specialist conditions

Many operational decisions depend on derived quantities rather than raw variables: wind chill and heat stress indices, icing and turbulence for aviation, significant wave height combined with period and direction for maritime work, illumination and moon phase, refraction and propagation conditions, and soil moisture for trafficability. Compute these from the appropriate source rather than approximating from surface variables. A good output is the specific derived parameter the decision requires with its method stated. Stop when the derived quantity is validated against observation where any is available.

Phase 8 — Communicate uncertainty honestly

Deliver the forecast with its uncertainty in a form that supports decision rather than defensiveness. Give the probability, the plausible range, the timing uncertainty and the specific scenario that would be worst for the operation. Avoid single deterministic numbers for anything beyond very short lead times, and avoid hedged language that transfers the decision back to the consumer. A good output is a briefing where the decision maker can state what they will do under each plausible outcome. Stop when the consumer understands the risk rather than merely the number.

Phase 9 — Monitor and update against triggers

Set automated monitoring against the defined thresholds with alerting, and update the assessment on a defined cadence and whenever a new model run materially changes the picture. Define in advance what change requires a re-brief, so updates are triggered rather than negotiated. For anticipatory action, agree the trigger and the response before the season, not during the event. A good output is a monitored trigger set with a documented update cadence and escalation route. Stop tuning when triggers fire early enough for the response to be executable.

Phase 10 — Record the forecast and verify

Archive what was forecast, when, with what confidence, and what actually happened, using official observations. Verification is what turns a forecast service into a capability, because it establishes real skill at each lead time for this location and this variable, and it identifies systematic bias that can then be corrected. Track false alarms and missed events separately, since they have different operational costs. A good output is a verification record with skill by lead time and variable. Stop only when the verification record is long enough to be statistically meaningful, then keep going.

Phase 11 — Support reconstruction and evidential use

Where conditions at a past time and place must be established for an investigation or a claim, use official archived observations or a documented reanalysis, obtained from the national meteorological authority with provenance or a certificate where available. State clearly whether the figure is an observation at a station, an interpolation, or a model reanalysis, and give the distance from the incident location. A good output is a conditions statement that an opposing expert can verify from the same public record. Stop when the provenance and the interpolation involved are explicit.

The platform ships this as a step-checked workflow in playbooks.php, so progress is recorded against a case rather than held in someone’s head.

Source register: what to collect from, and how

Sources are listed with their access model so you can plan around cost and licensing before you build a dependency on them. Open means no account required; registration means a free account or API key; licensed means paid or institutional access. Always confirm current terms — licensing changes, and a source that was free for research may not be free for commercial or evidential use.

Source	Access	What it gives you	How it is used here
ECMWF	Registration	European centre producing global deterministic and ensemble numerical weather prediction and the ERA reanalysis datasets.	Primary global forecast and ensemble source, plus the reanalysis used for reconstructing past conditions anywhere.
NOAA	Open	United States agency operating global and regional models, satellite and radar networks, and marine and climate observation systems.	Free access to global model output, marine observations and authoritative severe weather products and warnings.
World Meteorological Organization	Open	United Nations agency setting observation, coding and warning standards and coordinating national meteorological services.	Defines the observation and warning standards that make data from different national services comparable.
Met Office	Open	United Kingdom national meteorological service providing forecasts, warnings, archived observations and certified records.	Authoritative United Kingdom observations and certified historical records for evidential and claims use.
Copernicus Climate Data Store	Registration	Access platform for reanalysis, seasonal forecast and climate projection datasets including ERA5 hourly fields.	Retrieval of consistent global historical conditions for any location and time for reconstruction and baselines.
NOAA National Hurricane Center	Open	Official tropical cyclone forecasts, track and intensity guidance, advisories and historical best track data.	Authoritative tropical cyclone track and intensity forecasting for anticipatory action and evacuation triggers.
NOAA National Data Buoy Center	Open	Real-time and archived observations from moored buoys and coastal stations including wave, wind and sea temperature.	Ground truth for sea state and marine conditions used to validate model output against observation.
NOAA Tides and Currents	Open	Tide predictions, water level observations, currents and datums from coastal monitoring stations.	Tidal and water level data for coastal operations, storm surge assessment and flood reconstruction.
Aviation Weather Center	Open	Aviation meteorological products including routine and special reports, terminal forecasts, significant weather and icing guidance.	Standard coded observations and terminal forecasts used for aviation decisions and for point observations globally.
NOAA National Centers for Environmental Information	Open	Archive of climate and weather observations, storm event records and long-run climatological datasets.	Long climatological baselines and archived severe weather event records for reconstruction and risk assessment.
Copernicus Marine Service	Registration	Ocean analysis and forecast products covering currents, waves, sea ice, temperature and biogeochemistry.	Operational ocean forecasting for maritime planning where sea state and current drive the decision.
Open-Meteo	Open	Free interface serving multiple national model outputs and historical reanalysis through a simple programmatic interface.	Rapid programmatic access to multi-model forecasts and history for monitoring and automated threshold alerting.
Copernicus Emergency Management Service	Open	European service providing flood and fire early warning systems and rapid mapping activations for disasters.	Flood and fire warning products that provide the anticipatory action triggers used in humanitarian response.
USGS water data	Open	River gauge, groundwater and water quality observations across the United States with real-time and historical records.	Hydrological observation for flood assessment and for validating rainfall-driven flood forecasts against reality.

Prefer sources that publish a methodology and a revision history. A dataset that changes silently is a liability in any product that has to survive challenge.

Tooling

Tools commonly used against Meteorological Intelligence. None of these replace judgement, and each carries its own failure modes — know what a tool infers versus what it observes.

Model output retrieval and processing libraries — Pull gridded forecast fields programmatically and extract point series. Limitation: grid resolution does not represent a specific site without local correction.
Ensemble visualisation and postprocessing tools — Show forecast spread and threshold exceedance probability rather than a single value. Limitation: consumers routinely misread the ensemble mean as the forecast.
Geographic information systems — Combine forecast fields with terrain, assets and population for impact assessment. Limitation: overlaying coarse model output on fine geography implies false spatial precision.
Radar and satellite nowcasting viewers — Track precipitation and convection over the next few hours for immediate decisions. Limitation: skill decays rapidly beyond a few hours and coverage is uneven.
Reanalysis extraction tooling — Retrieve consistent historical conditions for any location and time for reconstruction. Limitation: reanalysis is a model product, not an observation, and must be labelled as such.
Threshold monitoring and alerting pipelines — Automate exceedance detection against defined operational thresholds. Limitation: requires calibrated site bias correction or it generates persistent false alarms.
Verification and skill scoring frameworks — Measure forecast performance by variable and lead time against observations. Limitation: needs a long record before scores are statistically meaningful.
Marine passage and wave analysis tools — Combine wave height, period and direction with vessel limits for maritime decisions. Limitation: significant wave height alone hides the period and direction that determine the actual risk.

AI skills and automation in detail

These are deterministic jobs with defined inputs and outputs, not open-ended prompting. Each is idempotent and cursor-based: interrupt one and it resumes where it stopped rather than duplicating work or losing progress.

Score Country Risk — Recomputes country risk from the weighted inputs and snapshots the result so movement over time is measurable.
Sync Intel Domains — Refreshes the reference and country-level intelligence datasets from their authorities.
Resolve Everything — Batch-resolves ASN, country, org and netblock for every IP from local reference datasets — no API calls, so it runs at millions of rows and works offline.
Summarise (Copilot) — Produces a narrative summary beside the underlying records. It explains; it never creates indicators or assigns attribution.
Generate Report — Assembles a sourced product from the current case or query, with provenance attached to each element.

A note on the boundary: the only skill that involves a language model is Summarise (Copilot), and it writes prose about records that already exist. Nothing else on this list involves generation of any kind. No indicator, relationship or attribution in the platform originates from a model. See the full skill list.

Tradecraft notes

The distinctions that separate a competent analyst from a fast one:

A forecast without a threshold is commentary. The discipline is converting model output into the probability that a specific operational limit is exceeded during a specific window, which is the only form a decision maker can act on.
Model grid output is not a site forecast. Terrain channelling, coastal effects, urban heat and valley cold pooling routinely produce differences larger than the forecast error, and quantifying the local bias is what makes the product usable.
Ensemble spread carries the decision-relevant information. Reporting the ensemble mean discards exactly the risk information the consumer needs, and a tight ensemble and a wide one with the same mean are entirely different situations.
Match the product to the lead time and stop pretending beyond it. Deterministic detail past about a week has no skill, and presenting it destroys credibility when it fails, taking the useful short-range products down with it.
Distinguish observation, forecast and reanalysis explicitly in every product. These are routinely conflated, and in evidential or claims work the difference between a station observation and a modelled value is the point that will be contested.
Verify continuously and publish the skill. Knowing that your wind forecast at this site is unbiased at two days and runs five knots low at four days is worth more than any model upgrade, and it is only obtainable by recording and checking.
Agree anticipatory action triggers before the season. Deciding what probability justifies prepositioning during the event guarantees late action, because the discussion always takes longer than the lead time available.
Significant wave height alone is not a sea state decision. Period and direction relative to heading determine whether a transfer or a landing is safe, and a modest height with a short steep period can be worse than a larger long swell.

Measuring whether it is working

Capability claims should be falsifiable. These are the measures that show whether work on Meteorological Intelligence is producing anything, and they are worth baselining before you change process or tooling.

Forecast skill by variable and lead time, verified against official observations at the specific locations supported, tracked over a rolling period.
False alarm rate and missed event rate against operational thresholds, tracked separately because their operational costs differ.
Lead time delivered before threshold exceedance, measured against the lead time the response actually requires to execute.
Proportion of decisions where the forecast product was used as intended, measured by decisions referencing a threshold probability rather than a single value.
Number of operational disruptions caused by conditions that were forecast but not acted on, which indicates a communication rather than a forecasting failure.
Share of site forecasts with a quantified local bias correction derived from station comparison rather than raw grid output.
Proportion of evidential conditions statements that correctly distinguish observation, interpolation and reanalysis with distance stated.

Beware of measuring volume alone. Indicator counts and report counts rise easily and say little; time-to-attribution, proportion of findings that survive review, and how often a product changed a decision say a great deal.

Common pitfalls

Treating a single deterministic run as fact while discarding the ensemble spread that quantifies how uncertain it actually is
Applying grid-cell output directly to a site, ignoring terrain, coastal and urban effects that dominate local conditions
Alerting on weather rather than on operational thresholds, so recipients quickly learn to ignore the notifications entirely
Missing the decision window, since a technically accurate warning issued after action becomes impossible has no value
Using unofficial aggregator apps for warnings when only the national authority product carries regulatory and insurance weight
Confusing forecast conditions with observed conditions when reconstructing an incident for claims or litigation

Legal and ethical considerations

Meteorological data from public agencies is generally free to use, but licence terms differ and some national services restrict commercial redistribution, while commercial providers impose strict reuse limits. Only official warnings from the responsible authority carry regulatory weight for duty-of-care, aviation and maritime compliance. Where forecasts inform safety decisions, document the data used and the reasoning, because liability follows the decision rather than the forecast. For incident reconstruction, prefer certified archived observations over model reanalysis if the record may be tested in court.

Data integrity: no fabrication, no drift, no hallucination

Intelligence that cannot be traced back to a source is not intelligence, it is assertion. Everything in this entry — and everything in the platform behind it — is built on a small number of non-negotiable rules.

Provenance on every record

Every indicator carries the source that supplied it, a first-seen and last-seen timestamp, and a sighting count. Where several feeds report the same artifact, each contribution is recorded separately rather than collapsed, so you can see whether a finding rests on one source or twelve. Source attribution travels with the data into every export, so a recipient can audit a claim without asking you for the working.

Nothing is invented to fill a gap

If the platform has no data for Meteorological Intelligence, it says so. Empty is displayed as empty — never padded with plausible-looking placeholder values, sample records or illustrative examples that a reader might mistake for observations. A dashboard with no rows is a true statement about collection coverage, and it is treated as a gap to close, not a blemish to hide.

Scoring is deterministic and reproducible

Threat scores, reputation grades and risk tiers are computed from stated inputs with fixed weights, not estimated. The same inputs always produce the same output, and the formula is visible rather than a black box. Aggregates are cached with an explicit time-to-live so a figure on screen is never silently stale — and when a heavy query exceeds its time budget the platform serves the last known-good value and labels it, rather than inventing a fresh number or hanging.

Where AI is used, and where it is not

Language models summarise and explain. They do not create indicators, assign attribution or manufacture relationships. No IP address, wallet, hash or identity in the platform originates from a model — every one is ingested from a named feed, resolved from a reference dataset, or entered by an analyst with a source recorded. Copilot output is presented as narrative alongside the underlying records, never in place of them, so a reader can always check the summary against the evidence.

Guarding against drift

Enrichment is additive and timestamped rather than overwriting. Reference data — sanctions lists, allocations, taxonomies — is re-synchronised from the authority on a schedule instead of being edited in place, so local copies cannot quietly diverge from the source of truth. Attribution is recorded with a confidence level and the reporting it rests on, and inferred relationships are labelled as inferred. When a source retracts or corrects, the correction propagates rather than leaving a stale assertion behind.

What this means for you

You can put a finding from this platform in front of a regulator, a court, a board or a partner agency and show where each element came from. That is the standard the tooling is built to — because in this work, being confidently wrong is more damaging than being usefully uncertain.

By the numbers

The taxonomy this entry belongs to is not a marketing list — it is the actual structure of the platform: 52 mission domains, 52 intelligence disciplines and 65 data points, each with a live dashboard behind it. Supporting that: 18 indicator types, 14 playbooks, 16 AI skills, 18 export formats and a 30-step automated pipeline.

This particular entry connects directly to 6 data points, 4 mission domains, 1 closely related entries — every one of them a tag you can follow, and a dashboard you can open.

Questions analysts actually ask

How far ahead is a forecast actually useful?

It depends entirely on the variable and the decision. Convective precipitation and thunderstorm timing have useful skill for hours; large-scale wind, temperature and pressure patterns retain skill to roughly a week; tropical cyclone track has useful skill to about five days while intensity is much weaker; and seasonal outlooks provide probabilistic signal about anomalies rather than specific conditions. The correct approach is to establish the skill for your variable, your location and your lead time through verification rather than accepting a general claim. Beyond the skill horizon, provide climatology and probability, not a forecast.

Deterministic or ensemble?

Ensemble for any decision with asymmetric consequences, which is most operational decisions. The deterministic run gives one plausible outcome; the ensemble gives the distribution, and the distribution is where the risk lives. A forecast of twenty five knots mean wind with a tight ensemble and the same mean with members spanning fifteen to forty knots are entirely different operational situations, and the deterministic run cannot distinguish them. Use the deterministic run for structure and timing detail, and the ensemble for the probability of exceeding your threshold, which is what the decision needs.

Can I use a consumer weather application for operational decisions?

For personal planning, yes; for operational or evidential purposes, no. Consumer applications present a single processed value with no indication of source model, run time, uncertainty or local bias, and the underlying source can change without notice. For operations, use the national meteorological service products and the model output directly, with your own site correction and thresholds. For evidential work, use official archived observations or a documented reanalysis with provenance from the meteorological authority, because a screenshot of an application is not admissible evidence of conditions.

What is reanalysis and when should I use it?

Reanalysis is a consistent reconstruction of past atmospheric and ocean conditions produced by running a fixed modern model over historical observations, giving complete global coverage on a regular grid for decades. Use it when you need conditions at a location and time with no nearby station, for climatological baselines, and for consistent long-run analysis. Do not use it as though it were an observation: it is a model product, it can differ substantially from a nearby station, and in evidential contexts you must label it as a modelled estimate and state the resolution and the nearest actual observation.

How should uncertainty be communicated to a commander or executive?

In decision terms with a probability and a worst case. State the probability that the operational threshold is exceeded during the window, the timing uncertainty, and the specific scenario that would be worst for the operation, then let them decide. Avoid single deterministic numbers, which invite false confidence, and avoid hedged narrative, which pushes the decision back to them without giving them the information to make it. The test of a good brief is whether the consumer can state what they will do under each plausible outcome before the event happens.

What weather evidence stands up in court?

Official archived observations from the national meteorological authority for the specific station, with a certificate or documented provenance, are strongest. Where the incident location is away from a station, a documented reanalysis or an interpolation is acceptable provided it is clearly labelled as modelled and the distance to the nearest observation is stated. Expect challenge on exactly that interpolation. Radar and satellite products are useful corroboration for precipitation and cloud. Consumer applications, unverified web archives and recollection are not evidence, and expert interpretation is normally required to put the material before the court.

How do we set anticipatory action triggers?

Before the season, in writing, with the response and the funding agreed at the same time. A trigger specifies the forecast product, the threshold, the probability level and the lead time, and it commits a defined action such as prepositioning, cash transfer or evacuation support. Set the probability level by weighing the cost of acting unnecessarily against the cost of acting late, which is a policy decision rather than a meteorological one. Then verify: record what triggered, what happened and whether the action was justified, and revise the thresholds on evidence rather than on the memory of the last event.

Standards, frameworks and further reading

Work that references a recognised framework is easier to defend, easier to hand over, and easier for a partner to consume:

WMO Technical Regulations and manuals on codes and observation, which govern how observations are made, coded and exchanged internationally.
ICAO Annex 3 on meteorological service for international air navigation, which governs aviation weather products and their formats.
WMO impact-based forecast and warning services guidance, which governs how warnings are expressed in terms of impact rather than parameters.
The Common Alerting Protocol, which governs the structured format for public warnings across hazards and authorities.
The Sendai Framework for Disaster Risk Reduction, which governs the early warning and anticipatory action approach used in humanitarian practice.
ISO 19115 geographic metadata, which governs provenance information accompanying gridded meteorological and oceanographic products.
IHO standards for hydrographic data and nautical charting, which govern the depth, tidal and datum data used in maritime planning.
WMO Global Data Processing and Forecasting System procedures, which govern the verification and exchange of numerical prediction products.

References

Primary sources and authoritative references for this entry. Publishers revise and retire material, so treat the retrieval date as part of the citation and re-check before relying on any of it in a formal product.

Numerical weather prediction and ERA reanalysis — European Centre for Medium-Range Weather Forecasts. Global deterministic, ensemble and reanalysis datasets
Weather, water and climate data services — National Oceanic and Atmospheric Administration. Model output, observations, warnings and marine data
Technical regulations and observation standards — World Meteorological Organization. International standards for observation, coding and warning services
Observations, forecasts and certified records — UK Met Office. Authoritative United Kingdom meteorological records and warnings
Climate Data Store — Copernicus. Reanalysis, seasonal forecast and climate projection datasets
Tropical cyclone forecasts and best track archive — NOAA National Hurricane Center. Official track and intensity guidance and historical cyclone records
Marine buoy observations — NOAA National Data Buoy Center. Real-time and archived wave, wind and sea temperature observations
Aviation meteorological products — Aviation Weather Center. Coded observations and terminal forecasts used for aviation decisions
Climate and weather archives — NOAA National Centers for Environmental Information. Long-run climatological records and archived severe weather events
Emergency Management Service early warning systems — Copernicus. European flood and fire early warning products supporting anticipatory action

Link integrity: every reference above was verified with a live request when this page was generated. Where a publisher had moved or withdrawn a document, the link was repointed at a preserved copy in the Internet Archive and marked as archived. Anything with no reachable copy anywhere had its link removed rather than left to rot — the source is still credited, it simply cannot be linked.

Put it into practice

The Quantus Intel threat intelligence platform operationalises this entry: alerts on site-specific thresholds with ensemble confidence and lead times matched to your response actions. Explore the platform, or browse the rest of the library by following any tag above.