METEOAPI · ENERGY
Client presentation · 2026
Hyperlocal weather intelligence

Weather is the
largest source
of uncertainty
in your portfolio.

Prepared for
Energy operators & traders
Region
Central & Eastern Europe
Operator
Meteo Centar d.o.o. · HR
meteoapi.eu
meteocentar@gmail.com
METEOAPI · ENERGY
02 · THE PROBLEM
The problem

Global forecast providers operate at 10–25 km resolution.
At that scale, orographic channelling, coastal sea-breeze, and valley shadow effects are invisible.

In Central Europe — where the Alps, Dinaric Alps, and Carpathians create complex local flow regimes — this is a structural problem. Across a portfolio of wind and solar assets, forecast errors compound into six- and seven-figure annual losses.

Source · Meteo Centar operational analysis 02 / 12
METEOAPI · ENERGY
03 · THE COST
What a forecast error actually costs

Small error, cubic pain.

Wind power scales with the cube of wind speed. Solar scales with cloud timing at minute resolution. Forecast errors don't scale linearly — they amplify.

10%
Solar irradiance error
A 10% error in next-day GHI forecast translates directly into imbalance-market costs.
15%
Power error from 5% wind error
Because P ∝ v³, a 5% wind-speed miss at hub height becomes a 15% deviation on the meter.
10–20%
Output swing per 1 m/s
A single meter-per-second error at hub height moves turbine output by 10–20% in the operating band.
P ∝ v³ · Cubic wind power law 03 / 12
METEOAPI · ENERGY
04 · RESOLUTION GAP
The resolution gap

Same terrain. Different reality.

A 10 km grid cell averages the wind farm on the ridge with the valley 8 km below. Our 1.5 km grid resolves individual valleys, lake-effect zones, and coastal gradients — the scale where weather actually matters for generation.

6.1 5.8 6.0 5.9 5.9 6.2 6.0 5.8 WIND FARM
Global provider
10 km grid · one value per cell
WIND FARM 9.4 m/s 4.1 m/s 6.3 m/s
MeteoAPI
1.5 km grid · 14 models blended
Ridge acceleration and lee-side drop both resolved at 1.5 km 04 / 12
METEOAPI · ENERGY
05 · OUR SOLUTION
Chapter 02 — the solution

1.5 km grid.
14 models blended.
Ready to consume.

A single JSON REST endpoint delivers irradiance components, hub-height wind, and grid-security parameters for any coordinate in our coverage area. No GRIB parsing. No log-law extrapolation. No separate irradiance decomposition library. Ready out of the box.

Coverage · AT · DE · SI · HR · BA · RS · CZ · SK · HU · RO · IT 05 / 12
METEOAPI · ENERGY
06 · SOLAR PV
Solar PV — full irradiance breakdown

GHI is not enough.
DNI, DHI, clearsky index — shipped.

For tracking arrays and tilted systems, a single GHI value is insufficient. DNI/DHI separation drives plane-of-array irradiance. Clearsky index is the standard ML feature for short-term PV forecasting. Both are derived server-side, delivered ready-to-ingest.

VariableAPI nameUnitWhy it matters for PV
Global horizontal irradiance · GHIshortwave_radiationW/m²Fixed-tilt PV generation baseline.
Direct normal irradiance · DNIdirect_normal_irradianceW/m²Tracker PV & CSP — the value that matters once panels follow the sun.
Diffuse horizontal irradiance · DHIdiffuse_horizontal_irradianceW/m²Low-light and overcast performance; plane-of-array calcs.
Clearsky GHIshortwave_radiation_clearskyW/m²Ineichen-Perez reference — enables clearsky-index nowcasting.
Clearsky indexclearsky_index0–1GHI / GHIcs — standard feature for PV forecast ML models.
Cloud cover (layered)cloud_cover_low/mid/high%PV ramp detection; overcast vs partial cloud discrimination.

DNI / DHI derived server-side via DIRINT decomposition (Perez et al. 1992). Clearsky GHI uses Ineichen-Perez with monthly Linke turbidity climatology from SoDa.

Integration · REST per farm → pvlib → generation forecast → exchange 06 / 12
METEOAPI · ENERGY
07 · WIND
Wind — native hub-height output

Not extrapolated.
Native at 80 / 120 / 180 m.

Most APIs ship 10 m wind and force operators to extrapolate via the log law — which breaks down in stable nocturnal conditions, exactly when wind farms generate hardest. Our WRF models emit wind on height-above-ground levels directly.

VariableAPI nameFleet / use
Wind @ 80 mwind_speed_80mV80-class, older fleet
Wind @ 120 mwind_speed_120mV126, N131 modern onshore
Wind @ 180 mwind_speed_180mTall onshore, offshore
Direction at hubwind_direction_*mTurbine yaw control
Shear exponent αwind_shear_exponentCustom hub extrapolation
Hub-height wind · schematic
10 m 80 m 120 m 180 m 18 km/h 34 km/h 47 km/h 61 km/h α = 0.21 shear exponent rotor
Native height-above-ground · no log-law extrapolation 07 / 12
METEOAPI · ENERGY
08 · GRID SECURITY
Beyond generation — operations & risk

Advance warning for unplanned outages.

Thunderstorms, high winds, and icing drive most weather-caused outages on the grid. MeteoAPI exposes the physical precursors so you can pre-position crews and curtail controllably — not reactively.

Storm risk
Thunderstorm probability
CAPE + wind shear + cold-front detection, recomputed every model run. Live risk for grid security and field-crew safety.
thunderstorm_probability
Structural
Wind gusts @ 10 m
Threshold triggers for turbine curtailment and overhead-line safety. Cross-referenced across all 14 ingested models.
wind_gusts_10m
Icing
Freezing level AGL
Ice accretion risk on transmission lines and turbine blades. Combined with humidity and temperature for icing probability.
freezing_level relative_humidity_2m
Convective
CAPE
Raw convective available potential energy — the standard severe-weather precursor for infrastructure risk.
cape
Roadmap

Webhook alerting — register thresholds, receive HTTP POST when exceeded

Planned · Enterprise
All probabilistic parameters available on Pro and Enterprise 08 / 12
METEOAPI · ENERGY
09 · USE CASES
Where clients put the API to work

Five proven use cases.

01

Solar farm generation forecasting

Hourly GHI/DNI/DHI + clearsky index per farm coordinate, feeding `pvlib` plane-of-array → DC/AC model → exchange submission. Multi-model spread surfaces forecast uncertainty for risk-adjusted dispatch.
shortwave_radiationclearsky_indexdirect_normal_irradiance
02

Wind farm power forecasting

Pull wind speed at the customer's exact hub height, apply turbine-specific power curve, output MW. The shipped shear exponent α extrapolates cleanly to non-standard hubs without the log-law's nighttime failure mode.
wind_speed_120mwind_shear_exponentair_density
03

Day-ahead market bidding

Long-horizon forecasts (up to 180 h) from our WRF ensemble suite, backed by ICON-EU and ARPEGE for cross-check. `temperature_2m` drives weather-normalised demand. Drop-in to existing bidding pipelines.
96 h horizontemperature_2mmulti-model cross-check
04

Grid security & extreme events

Thunderstorm probability, gusts, freezing level — pre-position maintenance crews, schedule controlled curtailment, and manage field-crew safety before events hit. Webhook alerting on the roadmap.
thunderstorm_probabilitywind_gusts_10mfreezing_level
05

Cross-country portfolio risk management

Uniform API across AT, DE, SI, HR, BiH, SR, MN, NMK, KO, CZ, SK, HU, IT, RO. Correlated-risk weeks — cold, calm, cloudy across the region — show up as ensemble spread in your Value-at-Risk pipeline, not as surprise imbalances.
ensemble spread → VaR inputportfolio-wide aggregationuniform variable names
All five reference integrations take < 1 sprint to stand up 09 / 12
METEOAPI · ENERGY
10 · INTEGRATION
Integration — one call, full stack

Hybrid site: PV + wind
in a single REST call.

Fetch full PV + wind variable set curl 
# Hybrid solar + wind site · Austrian Alps, 47.23 N / 15.67 E
curl "https://meteoapi.eu/api/v1/forecast" \
  --get \
  --data-urlencode "latitude=47.23" \
  --data-urlencode "longitude=15.67" \
  --data-urlencode "hourly=shortwave_radiation,direct_normal_irradiance,diffuse_horizontal_irradiance,\
                     shortwave_radiation_clearsky,clearsky_index,\
                     wind_speed_80m,wind_speed_120m,wind_speed_180m,\
                     wind_direction_120m,wind_shear_exponent,\
                     temperature_2m,cloud_cover,thunderstorm_probability" \
  -H "X-API-Key: met_your_key_here"
01 · Request
Coordinates + variables
One REST call per site per update cycle.
→
02 · Response
Column-oriented JSON
Ready for NumPy / Pandas / TSDB ingest. No GRIB parsing.
→
03 · Dispatch
Feeds pvlib + power curves
Plane-of-array + turbine curve → MW → exchange bid.
CORS enabled · X-RateLimit-Remaining on every response · OpenMeteo-compatible 10 / 12
METEOAPI · ENERGY
11 · VS GLOBAL PROVIDERS
How we compare

Built for energy.
Not adapted to it.

Factor Global providers · e.g. ECMWF API MeteoAPI.eu
Resolution in Alpine / Dinaric terrain 5–10 km 1.5–3 km
Models blended per forecast 1 Up to 14
Hub-height wind 100 m only (some) Native 80 / 120 / 180 m + shear exponent α
Irradiance components GHI only (most) GHI + DNI + DHI + clearsky GHI + clearsky index
Uncertainty signal Paid ensemble add-on Implicit from model spread — included
EU data residency Not guaranteed Yes · Croatia / EU only
Response format GRIB2 / NetCDF Clean JSON REST · CORS · OpenMeteo-compatible
Verified against ECMWF, Open-Meteo, and leading commercial weather APIs 11 / 12
METEOAPI · ENERGY
12 · NEXT STEPS
Where we go from here

A 30-day pilot.
One portfolio. Measurable delta.

We propose a structured pilot against your existing forecast provider:

  • 01 → You pick 5–10 representative sites from your current portfolio.
  • 02 → We run MeteoAPI in shadow mode alongside your incumbent for 30 days.
  • 03 → Joint readout: MAE / RMSE on hub-height wind and GHI, imbalance P&L delta.
  • 04 → Go / no-go on Enterprise contract, with pricing tied to the measured uplift.
General & trial
meteocentar@gmail.com
Enterprise / trading desk
Request demo call
API documentation
meteoapi.eu/docs
Service health
meteoapi.eu/api/health
Operated by
Meteo Centar d.o.o. · Croatia · EU
Thank you. METEOAPI · ENERGY SECTOR · 2026