Homeowners and utilities that depend on solar energy face a persistent question every time clouds roll in: do the panels on their roofs or in their fields stop producing electricity? The short answer is no. Solar modules continue to generate power under overcast skies because clouds scatter sunlight rather than block it entirely. The energy that reaches a tilted panel shifts from direct beam to diffuse radiation, and that diffuse component, while weaker, still drives measurable electrical output. How much output drops, and how engineers model that drop, shapes financial forecasts for solar installations across the cloudiest regions of the country.
Why diffuse irradiance keeps panels working under clouds
Sunlight arrives at Earth in two main forms. Direct beam radiation travels in a straight line from the sun to the surface. Diffuse radiation is sunlight that has been scattered by molecules, aerosols, and, most significantly, clouds. When an overcast layer covers the sky, the direct beam component can fall to near zero, but the diffuse component persists because clouds redirect photons across the entire sky dome rather than absorbing them all. That scattered light still carries energy, and photovoltaic cells convert whatever irradiance lands on their surface into electricity.
The U.S. Department of Energy has acknowledged this dynamic directly, noting that clouds reduce solar panel electricity production while also funding research into improving cell performance under low-light conditions. The agency’s Office of Science has supported work at national laboratories, including Brookhaven National Laboratory, aimed at designing materials that capture a broader spectrum of available light. The practical takeaway for system owners is straightforward: a cloudy day means less power, not no power.
How Perez and Sandia models quantify cloudy-day output
Two foundational modeling frameworks help engineers estimate exactly how much energy a tilted solar array collects when direct sunlight disappears. The first is the diffuse irradiance model developed by Perez and colleagues. Their simplified model for tilted surfaces, published in the journal Solar Energy, breaks the sky into discrete zones and assigns coefficients that describe how much scattered light reaches a panel at a given tilt and orientation. A companion paper by Perez et al. in 1990 extended this work by modeling daylight availability and irradiance components from direct and global irradiance, further refining predictions for overcast conditions.
The second framework comes from Sandia National Laboratories, whose PV Performance Modeling Collaborative maintains publicly available documentation of its sky diffuse method. This empirical approach estimates the sky-diffuse irradiance that falls on the plane of array, which is the actual surface a module presents to the sky. Sandia’s broader PV Array Performance Model treats total incident irradiance, the sum of direct beam plus diffuse plus ground-reflected light, as the input that determines electrical output. Because the model explicitly includes diffuse terms, it confirms that generation never drops to zero as long as some scattered light reaches the cells.
An open question is whether arrays tilted according to the Perez coefficients would capture meaningfully more diffuse energy per year than arrays set at a simple latitude tilt, especially in climates where diffuse irradiance accounts for a large share of total horizontal radiation. Running Sandia performance simulations against archived ground-station observations could test that hypothesis, but no published study has yet paired the two datasets in that specific way. The gap between modeled predictions and measured field performance under persistent cloud cover remains an area where additional validation would sharpen financial projections for solar projects in the Pacific Northwest, Great Lakes region, and similar overcast-heavy markets.
Ground measurements that confirm diffuse radiation persists
Models are only as reliable as the data used to calibrate them. NOAA’s Surface Radiation Budget Network, known as SURFRAD, operates ground stations across the United States that record solar radiation components, including diffuse irradiance measured with shaded pyranometers. These instruments physically block the direct solar beam so that only scattered sky radiation reaches the sensor. The network’s hourly observations, cataloged through the federal data repository, provide continuous evidence that measurable diffuse shortwave radiation reaches the surface even under heavy cloud cover.
What SURFRAD does not do, however, is co-locate its radiation sensors with operating photovoltaic arrays. That means researchers can confirm how much diffuse energy arrives at a given site but cannot directly pair those readings with real-time panel output from the same location. Bridging that gap would require either installing monitored PV systems at existing SURFRAD stations or matching SURFRAD timestamps with production data from nearby solar farms. Until that pairing exists, the connection between measured diffuse irradiance and actual kilowatt-hour production on cloudy days relies on model-based inference rather than direct observation.
What system owners and planners still need answered
Several practical questions remain open. First, no primary test results have applied the simplified Perez tilted-surface model to real-world portfolios of residential and commercial rooftops over multiple years. Many installers still default to rules of thumb, such as setting modules at or near the site’s latitude, instead of optimizing tilt for diffuse capture in overcast climates. Without comparative field data, it is difficult to quantify how much additional energy, if any, a Perez-based tilt strategy would deliver relative to simpler layouts once structural and aesthetic constraints are considered.
Second, there is limited public information on how accurately the Sandia sky-diffuse formulation predicts production on the very darkest days of winter in cloudy regions. Bankers and utility planners want to know not only annual energy yield but also the shape of the production profile during shoulder seasons and stormy weeks. If diffuse-focused models systematically overestimate output during prolonged overcast periods, financial pro formas could understate revenue risk. Conversely, if the models are conservative, developers may be leaving value on the table by undersizing inverters or storage relative to actual cloudy-day performance.
Third, the interaction between diffuse irradiance and modern module technologies is evolving. High-efficiency monocrystalline panels, bifacial designs that collect light from both sides, and anti-reflective coatings all respond differently to scattered light than older, lower-efficiency modules. Yet most widely used performance models were calibrated on earlier generations of equipment. Updating those models with field tests that isolate diffuse contributions would help clarify whether newer technologies narrow or widen the gap between sunny and cloudy-day output.
For homeowners, these unresolved issues translate into practical concerns: how much backup capacity is needed to ride through a string of gloomy days, how large a battery should be to smooth production swings, and whether investments in higher-efficiency panels are justified in regions where direct sun is rare. For utilities and grid operators, the same uncertainties affect resource adequacy planning and the design of incentive programs. If diffuse-driven production is more reliable than currently assumed, solar fleets may contribute more to winter peak coverage than conservative models suggest. If it is less reliable, planners may need to lean more heavily on complementary resources such as wind or demand response.
Answering these questions will require closer coordination between model developers, measurement networks, and system owners. One straightforward step would be to establish a handful of benchmark sites where SURFRAD-quality radiation sensors are installed alongside instrumented PV arrays using different module types and tilt strategies. Publishing anonymized production data from large utility-scale plants in consistently cloudy regions would also help researchers test and refine diffuse irradiance models. Over time, these efforts could turn the current patchwork of assumptions into a more robust, evidence-based understanding of how solar power behaves when the sun disappears behind the clouds.
Until that work is complete, the core message remains clear: clouds dim solar panels, but they do not turn them off. Diffuse irradiance ensures that even under a solid gray sky, photons continue to reach the silicon, and electrons continue to flow. For homeowners and planners alike, recognizing the value of that scattered light is essential to designing systems that perform reliably in the real, often cloudy, atmosphere rather than in an idealized world of permanent blue skies.
More from Morning Overview
*This article was researched with the help of AI, with human editors creating the final content.
