What Is GHI (Global Horizontal Irradiance)? Definition & Guide

Key Takeaways

Global horizontal irradiance (GHI) is the total solar energy striking a flat, horizontal surface — it combines direct beam radiation (DNI) and diffuse scattered radiation (DHI) into a single value
GHI is the standard metric for comparing solar resource between locations, with values ranging from about 800 kWh/m²/year in northern Europe to over 2,500 kWh/m²/year in desert regions
The formula GHI = DNI × cos(θz) + DHI relates all three irradiance components, where θz is the solar zenith angle
GHI data sources include ground-based pyranometer stations, satellite-derived models (NREL NSRDB, NASA POWER, Solargis), and Typical Meteorological Year (TMY) datasets
GHI alone is not sufficient for accurate PV system design — solar design software converts GHI to plane-of-array (POA) irradiance using transposition models that account for panel tilt and azimuth
Peak sun hours per day equal GHI in kWh/m²/day. A site with GHI of 5.0 kWh/m²/day receives five peak sun hours — the equivalent of five hours at 1,000 W/m²

What Is Global Horizontal Irradiance (GHI)?

Global horizontal irradiance (GHI) is the total amount of solar radiation received per unit area on a horizontal surface at ground level. It includes both the direct beam component from the sun (projected onto the horizontal plane) and the diffuse component scattered by the atmosphere from all sky directions.

GHI is the most widely used metric for assessing solar resource potential. When someone says a location receives “5 kWh/m²/day of solar radiation,” they are referring to GHI. Every major solar radiation database — NREL NSRDB, NASA POWER, Meteonorm, Solargis — reports GHI as the primary output.

GHI is the starting point for all solar energy calculations. A site’s annual GHI tells you the theoretical maximum energy available on a flat surface. But real PV systems are tilted toward the sun, which means GHI must be converted to plane-of-array (POA) irradiance before it can predict actual system output. In high-latitude locations, tilted panels can receive 10–25% more annual irradiance than horizontal GHI suggests. This conversion, handled by transposition models in solar design software, is why GHI alone never tells the full story.

Types of GHI Data and Measurement

GHI data comes from several measurement and modeling approaches, each with different accuracy, coverage, and use cases:

GHI from Ground Stations (Pyranometer)

Pyranometers are dome-shaped sensors that measure total solar radiation on a horizontal surface. Research-grade instruments (ISO 9060 Class A) achieve 2–3% annual measurement uncertainty. Networks like the Baseline Surface Radiation Network (BSRN) and national weather services maintain thousands of stations worldwide. Ground-measured GHI is the gold standard for validating satellite models, but station coverage is sparse — most project sites are 50–200 km from the nearest station.

Satellite-Derived GHI

Geostationary satellites (GOES, Meteosat, Himawari) image cloud cover every 5–30 minutes. Physical models convert cloud observations and atmospheric data into surface GHI estimates with 4–6% annual uncertainty. NREL’s NSRDB covers the Americas at 4 km × 4 km spatial resolution and 30-minute temporal resolution. Solargis and SolarAnywhere offer global coverage. Satellite data is the default input for most solar project feasibility studies.

TMY GHI Data

Typical Meteorological Year (TMY) datasets compile representative months from 15–30 years of historical records into a single synthetic year. TMY GHI represents long-term average conditions, filtering out anomalous years. PVsyst, SAM, and other simulation tools use TMY files (in TMY3 or EPW format) as their standard weather input. TMY data is appropriate for long-term yield estimates but cannot capture specific year variability needed for P90 analysis.

Real-Time GHI Measurement

On-site pyranometers or reference cells provide real-time GHI readings for operating solar plants. This measured GHI is compared against actual system output to calculate the performance ratio (PR) — a key indicator of system health. Real-time GHI also feeds into nowcasting models that predict short-term solar output for grid operators. Monitoring-grade sensors cost $200–$2,000 and require periodic recalibration every 1–2 years.

GHI by Location

GHI varies significantly with latitude, climate, altitude, and cloud cover. The table below shows representative values for major solar markets:

Location	Annual GHI (kWh/m²/yr)	Peak Sun Hours/day	Solar Rating	Typical System Yield (kWh/kWp)
Phoenix, Arizona	2,350	6.4	Excellent	1,750–1,900
Atacama Desert, Chile	2,500	6.8	Excellent	1,850–2,050
Rajasthan, India	2,100	5.8	Very Good	1,600–1,750
Southern Spain	1,900	5.2	Very Good	1,500–1,650
Rome, Italy	1,750	4.8	Good	1,350–1,500
Sydney, Australia	1,800	4.9	Good	1,400–1,550
Munich, Germany	1,200	3.3	Moderate	950–1,100
London, UK	1,000	2.7	Low	800–950
Amsterdam, Netherlands	1,050	2.9	Low	850–1,000
Oslo, Norway	900	2.5	Low	750–900

Peak sun hours per day equals annual GHI divided by 365. A location with 1,825 kWh/m²/year of GHI receives 5.0 peak sun hours per day — meaning the total daily energy equals what five hours of full 1,000 W/m² sunshine would deliver.

System yield (kWh/kWp) is always lower than GHI due to module efficiency losses, temperature derating, inverter losses, soiling, shading, and wiring losses. The ratio of actual yield to GHI-based theoretical yield is the performance ratio (PR), typically 75–85% for well-designed systems.

The GHI Formula

GHI is defined by a straightforward relationship between the three primary irradiance components:

Global Horizontal Irradiance Formula

GHI = DNI × cos(θz) + DHI

Where:

GHI = Global Horizontal Irradiance (W/m² instantaneous, or kWh/m²/day for daily totals)
DNI = Direct Normal Irradiance — the beam radiation measured perpendicular to the sun’s rays
θz = Solar zenith angle — the angle between the sun and the vertical (directly overhead). At solar noon in summer at the equator, θz approaches 0° and cos(θz) approaches 1.0
DHI = Diffuse Horizontal Irradiance — radiation scattered by the atmosphere arriving from all sky directions

When the sun is directly overhead (θz = 0°), cos(θz) = 1.0 and the full DNI value contributes to GHI. As the sun drops toward the horizon (θz increases toward 90°), cos(θz) shrinks toward zero and the DNI contribution to GHI diminishes. At sunrise and sunset, GHI is almost entirely composed of DHI.

On a clear day at solar noon in a desert location, typical values might be: DNI = 950 W/m², θz = 20°, DHI = 100 W/m². That gives GHI = 950 × cos(20°) + 100 = 950 × 0.94 + 100 = 993 W/m². On a fully overcast day, DNI drops to zero and GHI equals DHI — perhaps 150–300 W/m² depending on cloud thickness.

GHI vs. POA Irradiance — Why the Distinction Matters

GHI measures radiation on a flat, horizontal surface. But solar panels are rarely horizontal — they are tilted toward the equator (or mounted on trackers) to capture more direct sunlight. The irradiance actually reaching a tilted panel is called plane-of-array (POA) irradiance, and it is always different from GHI. At high latitudes, a properly tilted panel receives 15–25% more annual irradiance than horizontal GHI. Transposition models (Perez, Hay-Davies, isotropic) convert GHI, DNI, and DHI into POA irradiance by accounting for tilt angle, azimuth, ground reflectance (albedo), and circumsolar brightening. This is why accurate solar design software needs all three irradiance components — not just GHI — to produce reliable energy yield predictions.

Using GHI in Solar Design

Site Assessment and Feasibility

GHI is the first number to check when evaluating a potential solar project site. Annual GHI above 1,500 kWh/m²/year generally indicates a strong solar resource. Below 1,000 kWh/m²/year, projects are still viable but require lower system costs or higher electricity rates to achieve acceptable returns. Compare GHI values from multiple data sources — if NSRDB and Solargis disagree by more than 5%, investigate which source has better validation for that region.

System Sizing

A quick system sizing estimate uses GHI directly: divide the annual energy target (kWh) by the GHI (kWh/m²/year) and the expected performance ratio (typically 0.78–0.82 for residential, 0.80–0.85 for ground-mount). The result is the required array capacity in kWp. For example, a home needing 10,000 kWh/year in Munich (GHI = 1,200 kWh/m²/year) with PR of 0.80 requires 10,000 / (1,200 × 0.80) = 10.4 kWp.

Shading Impact on GHI Components

When an obstruction blocks the sun, the DNI component of GHI drops to zero for the shaded area while the DHI component continues at a reduced level depending on how much sky the obstruction blocks. Shadow analysis software models this separation hourly throughout the year, calculating separate beam and diffuse shading losses. In overcast climates where DHI is 50–60% of GHI, shading losses from nearby objects are less severe than in clear-sky climates where DNI dominates.

GHI Data Quality Checks

Before using any GHI dataset in a simulation, verify these basics:

Check the data period. At least 10 years of data is recommended for TMY construction. Short records (under 5 years) can produce GHI values biased by anomalous weather patterns.
Compare multiple sources. Cross-reference NSRDB, NASA POWER, and Solargis for the same coordinates. Consistent results (within 3–4%) increase confidence.
Verify the spatial resolution. Coarse-resolution data (50 km grid) can miss local effects like coastal fog, mountain shadows, or urban heat islands. Use the highest resolution available for the project location.
Review the decomposition model. When only GHI is measured (as at most pyranometer stations), DNI and DHI must be estimated using decomposition models (Erbs, Orgill-Hollands, DISC). Different models produce different DNI/DHI splits, which affects tilted-surface irradiance by 2–5%.
Account for inter-annual variability. GHI varies 2–5% year-to-year depending on location. Use the full historical record to quantify P50, P75, and P90 GHI values for financial modeling.

Pro Tip

When comparing GHI values between databases, check whether the reported values use a satellite-only model or a satellite-plus-station-blended model. Blended products (like NSRDB’s PSM v3) adjust satellite estimates using nearby ground station measurements, reducing annual bias to under 3%. Pure satellite models can carry 5–8% bias in regions with complex terrain or persistent aerosol loading.

Access Site-Specific Irradiance Data in Your Design

SurgePV integrates high-resolution GHI, DNI, and DHI databases to deliver accurate energy yield simulations for any location, tilt angle, or tracking configuration.

Book a Demo

No commitment required · 20 minutes · Live project walkthrough

GHI Maps and Regional Patterns

Global GHI follows predictable geographic patterns driven by latitude, cloud cover, and atmospheric clarity:

Highest GHI (2,200–2,600 kWh/m²/year): Arid regions between 15°–35° latitude on both hemispheres. The Atacama Desert, Saharan Africa, Arabian Peninsula, and Australian outback receive the strongest solar resource due to clear skies, low humidity, and high elevation in some areas.
High GHI (1,700–2,200 kWh/m²/year): Mediterranean climates, semi-arid regions, and subtropical zones with distinct dry seasons. Southern Europe, the US Southwest, northern India, and much of southern Africa fall in this range.
Moderate GHI (1,200–1,700 kWh/m²/year): Mid-latitude temperate zones with significant seasonal variation. Central Europe, the northeastern US, Japan, and southern Australia receive adequate solar resource for cost-effective PV.
Low GHI (800–1,200 kWh/m²/year): Northern Europe, the Pacific Northwest, and regions with persistent cloud cover. Solar is still deployed in these areas — Germany at ~1,100 kWh/m²/year GHI is the world’s fourth-largest solar market — but financial returns depend more on retail electricity prices and incentive structures.

Sources

NREL National Solar Radiation Database (NSRDB) — High-resolution GHI, DNI, and DHI data for the Americas at 4 km spatial and 30-minute temporal resolution
NASA POWER (Prediction of Worldwide Energy Resources) — Global solar irradiance data derived from satellite observations, covering all land surfaces at 0.5° resolution
Meteonorm — Global irradiance database combining station measurements and satellite data with interpolation for any location worldwide

Frequently Asked Questions

What is GHI in solar energy?

GHI (Global Horizontal Irradiance) is the total solar radiation measured on a flat, horizontal surface at ground level. It combines the direct beam component from the sun (DNI projected onto the horizontal plane) and the diffuse component scattered by the atmosphere (DHI). GHI is expressed in kWh/m²/day or kWh/m²/year and is the standard metric used to compare solar resource potential between locations. Typical values range from about 2.5 kWh/m²/day in northern Europe to 7.0 kWh/m²/day in desert regions.

What is a good GHI value in kWh/m²/day for solar panels?

A GHI of 4.0–5.0 kWh/m²/day (1,460–1,825 kWh/m²/year) is considered good for solar panel installations, producing typical system yields of 1,200–1,500 kWh/kWp/year. Values above 5.5 kWh/m²/day are excellent and common in the US Sun Belt, Middle East, and Australia. Values below 3.0 kWh/m²/day (under 1,100 kWh/m²/year) are on the lower end but still support viable solar projects in countries like Germany and the UK, where high electricity prices and feed-in tariffs compensate for the lower resource. The actual energy output depends on panel tilt, orientation, shading, and system losses — all factors that solar design software accounts for when converting GHI into a realistic yield estimate.

How is GHI different from POA irradiance?

GHI measures radiation arriving on a horizontal surface. POA (plane-of-array) irradiance measures radiation arriving on the actual tilted surface of a solar panel. Since panels are typically tilted toward the equator at an angle close to the site’s latitude, POA irradiance is usually higher than GHI — by 10–25% at mid and high latitudes. The conversion from GHI to POA requires transposition models that use all three irradiance components (GHI, DNI, DHI) along with panel tilt, azimuth, and ground albedo. This is why solar simulation tools need the full irradiance dataset, not just GHI, to predict accurate energy production. Shadow analysis software further adjusts POA irradiance by calculating how obstructions reduce both beam and diffuse components at different times of day and year.