In this document
We will explain how variability in solar radiation — across timescales ranging from minutes to decades — directly impacts energy production and system performance. Understanding these fluctuations helps design systems that meet energy production targets, optimize financial returns, and ensure stable grid integration.
Usage in Solargis platform |
|---|
This approach is used in Solargis Prospect and Solargis Evaluate. |
Overview
Solar radiation at any given location is never constant. It fluctuates across multiple timescales, each with distinct causes and implications for solar project design, financial modeling, and operational planning. Solargis Time Series data is the foundational source of information for all variability analyses described in this document, providing high-resolution historical solar radiation records that enable statistical analysis at every scale — from sub-hourly fluctuations to multi-decadal trends. Time Series data is available in Solargis Evaluate.
Statistical analysis of solar radiation data provides a detailed understanding of solar energy availability, helping solar project stakeholders make informed decisions about site selection, system design, and technology choices. These insights enable more accurate financial models and performance predictions over the lifecycle of solar power plants.
This document covers the following variability types:
Very short-term to short-term variability (1-minute to 60-minute): rapid fluctuations caused by passing clouds and atmospheric changes, critical for grid integration and storage sizing.
Intraday variability: day-by-day differences in daily solar radiation totals within a month or season, relevant for operational planning and short-term forecasting.
Seasonal variability: systematic changes in solar radiation between months and seasons driven by the Earth's orbital geometry, essential for system sizing and yield prediction.
Interannual variability: year-to-year deviations from the long-term average, driven by weather cycles and climate variability, critical for financial risk assessment.
Long-term cycles and trends: gradual changes in solar radiation over decades, including the influence of atmospheric aerosols, land-use change, and climate-driven shifts, relevant for long-term asset management and portfolio planning.
Very short-term to short-term variability (1-minute to 60-minute)
Very short-term and short-term variability refers to rapid fluctuations in solar irradiance occurring on timescales from one minute up to one hour. These fluctuations are primarily driven by cloud passage, aerosol events, and rapidly changing atmospheric conditions.
What it is
At the sub-hourly level, solar irradiance can change dramatically within seconds as clouds move across the sky. The magnitude and frequency of these fluctuations depend on local climate conditions — sites with frequent convective cloud activity or coastal fog experience more intense variability than arid or semi-arid locations.
Analysis of very short-term fluctuations is typically performed using histograms and frequency distribution charts. This approach provides a statistical view of how solar irradiance is distributed across a range of values over time, revealing the most common irradiance levels and the probability of encountering extreme drops or spikes.
Moreover, these fluctuations can be analyzed by looking at the ramps in solar radiation data, i.e. the rate at which the radiation increases or decreases. Solargis has published research on the ramps in Global horizontal irradiance (GHI) and on variability of PV power output (PVOUT). The GHI ramps are provided in Solargis Prospect as a map layer.
When and why to use it
Understanding sub-hourly variability is essential for:
Grid integration: Sudden drops or spikes in solar power output can destabilize grid frequency and voltage. Knowing the frequency and magnitude of such events allows grid operators and system designers to define appropriate ramp-rate controls and ancillary service requirements.
Curtailment planning: Identifying how often and by how much generation may exceed grid capacity enables better planning of curtailment strategies to minimize energy losses.
Inverter and power electronics design: Frequent large fluctuations impose mechanical and thermal stress on inverters and other power electronics; short-term variability data helps define design margins.
Note: Solargis Time Series data at 15-minute and 10-minute resolution provides the foundation for short-term variability analysis.
Intraday variability
Intraday variability describes the day-by-day differences in the shape of the radiation profile and the daily sum of solar radiation within a given period — typically within a month or season. Unlike interannual variability, which compares full annual totals, intraday variability focuses on how much each day's solar profile deviates from the typical daily pattern.
What it is
The daily profile aggregates the average irradiance at each hour of the day over all days in a specified period, typically a month, revealing the characteristic daily curve. Deviations from this curve, caused by cloud cover, fog, or weather fronts, represent intraday variability.
Analyzing the average hourly and sub-hourly solar radiation throughout the day for each month helps understand the typical daily solar radiation patterns across different months. It offers a precise view of how solar irradiance varies throughout the day and year, which is essential for optimizing system performance and ensuring that energy production matches power delivery commitments on a daily basis.
When and why to use it
Understanding intraday variability is important for:
Energy storage sizing: Daily profiles support the sizing of energy storage systems, ensuring enough capacity to store excess energy during peak sunlight hours and discharge it when irradiance is low.
Demand matching: Knowing how reliably solar generation follows the expected daily profile each day of the month helps assess how well generation aligns with delivery schedules.
Maintenance planning: Identifying months or seasons with higher day-to-day variability helps schedule maintenance activities during periods of more predictable, lower-risk output.
Short-term forecast validation: Day-by-day profile comparisons serve as a baseline for validating short-term solar forecasts.
Note: Daily profiles are directly available as part of the statistical outputs from Solargis Time Series in Solargis Evaluate.
Seasonal variability
Seasonal variability describes the systematic and predictable changes in solar radiation between months and seasons throughout the year. It is primarily driven by the Earth's axial tilt and orbital geometry, which cause the sun's elevation angle and day length to vary with season.
What it is
Monthly and annual solar radiation totals — sums, averages, minima, and maxima — are the primary tools for characterizing seasonal variability. Yearly and monthly irradiation sums are essential for creating accurate financial models, including revenue projections, return on investment (ROI), and payback periods. Monthly totals also facilitate comparisons across different locations or years, helping to identify the best-performing sites and assess long-term trends in solar resource availability.
Minimum and maximum irradiation sums reveal the range of solar radiation values, highlighting the variability of solar energy availability and helping developers assess the predictability of solar resources, which is crucial for financial modeling and risk assessment.
Understanding the extremes of operating conditions (both low and high) is also required for designing systems to handle worst-case scenarios, ensuring reliability and preventing over- or under-sizing. Identifying periods with potential low irradiance enables planning of maintenance schedules or storage needs.
Simplified seasonality indices, which indicate the magnitude of the seasonal changes, are available is Solargis Prospect.
When and why to use it
Seasonal variability analysis is foundational for:
System sizing and layout optimization: Knowing the seasonal distribution of solar radiation is critical for optimizing panel tilt angle, tracker strategies, and inverter sizing to maximize annual yield.
Financial modeling: Monthly and annual irradiation sums feed directly into energy yield calculations, revenue forecasts, and payback period analyses.
Grid planning: Understanding seasonal generation profiles helps grid operators plan capacity reserves and seasonal balancing requirements.
Power delivery contract planning: Sites with pronounced seasonal variation may need to account for the variability in their power delivery contract, or purchase additional power to balance the seasons with lower production.
Tip: Always review both minimum and maximum monthly values — not just the average — to capture the full range of seasonal performance risk at a site.
Interannual variability
Interannual variability quantifies the year-to-year fluctuations in solar radiation at a specific location, typically expressed as a percentage of the long-term average. These fluctuations result from natural weather cycles and stochastic variations, causing annual solar radiation to deviate by a few percent from the long-term average.
What it is
Understanding interannual variability is essential for developers and investors to assess the reliability and predictability of solar resources over the long term.
For a specific solar project, interannual variability is calculated using the historical series of annual solar radiation or PV output values. While variability is often calculated for a single year, the analysis can consider longer period to reflect the balancing effect of multiple years. In general, the longer the considered period, the lower the variability of that period as a whole with respect to the long-term average.
Assuming that interannual variability follows a normal distribution, the following steps are commonly used:
Calculate the standard deviation of annual values over the whole available data period.
Divide the standard deviation by the square root of the number of years in the period considered for the interannual variability.
Divide the result by the average value of the whole dataset to express it as a percentage.
Convert the value to any desired P-value confidence level, such as P90, by multiplying it by the appropriate factor (e.g., 1.282 for P90).
The standard deviation of the Solargis Time Series data in yearly summaries is provided as the long-term variability map layer in Solargis Prospect.
Interannual variability can be calculated for solar radiation or expected PV output. For PV output, the calculation follows the same steps. However, a simplified approach assumes that the interannual variability of PV output is identical to that of solar radiation. This approximation has limitations, as it neglects the effects of factors such as cell temperature, system losses, and non-linearities caused by partial shading of modules.
Important: The simplified assumption that PV output variability equals solar radiation variability may underestimate real PV output variability in climates with significant temperature swings or complex shading conditions.
GHI interannual variability — example values
The table below illustrates how GHI interannual variability decreases as the considered period increases..
Nearby city | Country | Variability [%] 1 year | Variability [%] 5 years | Variability [%] 10 years | Variability [%] 25 years |
|---|---|---|---|---|---|
Kosice | Slovakia | 3.8 | 1.7 | 0.5 | 0.1 |
Fresno | United States | 2.5 | 1.1 | 0.4 | 0.1 |
Kurnool | India | 2.3 | 1.0 | 0.3 | 0.1 |
Calama | Chile | 1.3 | 0.6 | 0.2 | 0.0 |
Upington | South Africa | 1.3 | 0.6 | 0.2 | 0.0 |
Table of GHI interannual variability of a period of 1, 5, 10, and 25 years for several sample sites
When and why to use it
P-value energy yield estimates (P50, P90): Interannual variability is a key input into the calculation of uncertainty of a solar resource data, which in turn drives probabilistic energy yield calculations. P90 estimates, for example, indicate the annual energy production level that is exceeded in 9 out of 10 years — an important metric for project debt financing.
Financial risk assessment: Lenders and investors use interannual variability to quantify downside energy production risk and size debt service reserve accounts accordingly.
Long-term performance prediction: Multi-year variability analysis helps assess whether a site's resource is stable enough to support long-term power purchase agreements (PPAs).
Long-term cycles and trends
Long-term variability describes gradual shifts in solar radiation over decades. Unlike shorter-term variability, these changes are not simply noise around a stable mean — they represent directional trends or multi-decadal cycles driven by large-scale atmospheric and climatic processes.
What it is
Research published in Remote Sensing of Environment (Past, current and future solar radiation trends in Europe) provides evidence on multi-decadal changes in solar radiation in Europe. The study analyzed surface solar radiation trends using long-term ground measurements and satellite-based models to describe observable trends. Due to availability of high-quality long-term ground measurements solely in Europe, it focused on Europe only.
The study concluded that Europe experienced a period of “brightening” - increase in solar radiation of +3.1 W/m2 per decade between 1994 and 2023. However, it also notes that this trend will not continue in the coming decades. Positive and negative trends are described in the published literature for several other global regions. For solar power plants with 25–40 year operational lifetimes, ignoring such a long-term trend in solar resource could lead to systematic over- or under-estimation of energy production in later project years.
The quality of the data and the data processing methods based on which the trend claims are made vary greatly between the studies. Furthermore, trends observed in the past decades are not guaranteed to continue in the future. For this reason, analyses and forecasts which integrate trends into PV yield estimates should be made carefully with a proper understanding of the limitation of the data.
When and why to use it
Long-term trend analysis is relevant for:
Long-term asset management: Portfolio managers and asset owners need to account for potential systematic shifts in solar resource when projecting revenues and asset values over multi-decade horizons.
Refinancing and repowering decisions: Long-term trends can affect the solar resource baseline used in refinancing assessments or decisions about extending a plant's operational life.
Climate risk disclosure: Investors and lenders increasingly require assessment of physical climate risks, including changes in solar resource availability, as part of ESG and TCFD-aligned reporting frameworks.
Model calibration and updating: Awareness of long-term trends informs how often historical solar radiation datasets should be updated and whether recent years should be weighted more heavily in long-term average calculations.
Note: Solargis continuously updates its historical solar radiation archive to incorporate the latest satellite data and atmospheric inputs, ensuring that long-term averages and trend analyses reflect the most current observational record available.