How Much Do Solar Panels Produce on Cloudy and Rainy Days in Los Angeles? Measured
Solar panels keep producing under clouds, but how much depends on how thick the clouds are. Across 219 rooftop systems and 241,662 system-days in the Los Angeles area, a partly cloudy day gave about 70% of what the same roof makes on a typical clear day, a mostly cloudy day about 44%, and a dark, overcast day about 23%. Light rain left about 63%, heavy rain about 23%. Over a whole year, clouds and rain cost LA roofs roughly 14% of what an always-clear year would give.
Output by sky condition
Each day sorted by how much sunlight reached the ground, independent of the panels; output compared with the same roof on a typical clear day.
| Day type | Sunlight that day, % of a clear day | Output, % of a clear day (median) | Most days fall in* | Days a year** | Share of a year’s output** |
|---|---|---|---|---|---|
| Clear | 90% or more | 100% | 84–100% | 194 | 61.0% |
| Mostly sunny | 75–90% | 90% | 61–100% | 100 | 28.1% |
| Partly cloudy | 55–75% | 70% | 38–95% | 41 | 7.7% |
| Mostly cloudy | 35–55% | 44% | 24–71% | 16 | 2.1% |
| Dark overcast | under 35% | 23% | 9–45% | 15 | 1.2% |
| *Middle 80% of system-days (p10–p90), capped at 100%. **San Fernando Valley weather, average of 2023–2025. Based on 241,662 system-days. | |||||
“Sunlight that day” is the solar energy reaching the ground in that area, from the ERA5 weather record, as a share of a clear day at the same place and time of year. It sorts days by the weather alone, so the panels’ own output does not decide which bin a day falls into. Within each bin, days vary a lot: a morning of marine layer that burns off by noon and a day of thin high cloud can both land in “mostly sunny,” and panels respond differently to each.
Why panels still work under clouds: they convert all daylight, not only direct sun. On a gray day most of the light is diffuse — scattered by the clouds — and panels keep producing from it, just less. The physics is explained in do solar panels work in cloudy or rainy weather.
Rainy days
Daily rainfall at the site’s nearest weather point.
| Rain that day | Output (median) | Most days fall in* | Days a year** |
|---|---|---|---|
| Light rain (1–5 mm) | 63% | 28–96% | 21 |
| Moderate rain (5–20 mm) | 46% | 19–86% | 15 |
| Heavy rain (20 mm or more) | 23% | 7–60% | 10 |
| *Middle 80% of system-days. **San Fernando Valley, average of 2023–2025. Rainfall from ERA5 reanalysis. | |||
A day with light showers still gave about 63% of clear-day output: in Los Angeles, rain often falls in a few hours with sun before or after. Heavy-rain days, usually under a solid storm system, dropped to about 23%. The San Fernando Valley point averaged 46 days a year with at least 1 mm of rain in 2023–2025.
Month by month: how much the weather takes
Average output as a share of an all-clear month, and how many days were clear.
| Month | Clear days (San Fernando Valley) | Average day, % of a clear day |
|---|---|---|
| January | 38% | 85% |
| February | 28% | 82% |
| March | 29% | 85% |
| April | 60% | 88% |
| May | 48% | 83% |
| June | 71% | 90% |
| July | 89% | 96% |
| August | 81% | 94% |
| September | 49% | 87% |
| October | 52% | 90% |
| November | 41% | 85% |
| December | 51% | 81% |
| Clear days: sunlight 90% or more of a clear day (ERA5), 2023–2025. Output: mean over all valid system-days in that month, 2019–2026. | ||
July is the clearest month: 89% of days were clear and roofs averaged 96% of their clear-day output. The weather takes the most in February, May, December, when roofs averaged 81–83% of clear-day output. May stands out among the warm months (83%): that is the season of morning low clouds, “May Gray.” Winter months lose output twice: shorter, lower sun on every day, and more cloudy days. See how much production drops in winter.
| Day type | San Fernando Valley | Santa Clarita Valley | Los Angeles basin | Ventura County |
|---|---|---|---|---|
| Mostly sunny | 91% | 94% | 87% | 91% |
| Partly cloudy | 69% | 67% | 72% | 68% |
| Mostly cloudy | 45% | 42% | 44% | 43% |
| Dark overcast | 23% | 23% | 23% | 25% |
| Partly cloudy or worse, days a year | 72 | 58 | 76 | 66 |
| Clear days a year | 194 | 208 | 191 | 177 |
| Each area uses its own nearest weather point. — = fewer than 300 system-days. Days a year: average of 2023–2025. | ||||
On a cloudy day of the same thickness, panels in the valleys, the basin and Ventura County lost about the same share. Where areas differ is in how many such days they get: Santa Clarita Valley averaged 58 partly-cloudy-or-worse days a year in 2023–2025, Los Angeles basin 76.
The biggest storms of 2023–2025
The wettest days at the San Fernando Valley weather point, and the two days after.
| Storm day | Rain | That day | Next day | Day after | Systems |
|---|---|---|---|---|---|
| February 24, 2023 | 82 mm | 11% | 30% | 65% | 85 |
| December 24, 2025 | 77 mm | 5% | 60% | 24% | 195 |
| January 9, 2023 | 72 mm | 11% | 37% | 80% | 79 |
| February 4, 2024 February 2024 atmospheric river | 68 mm | 21% | 6% | 19% | 158 |
| February 5, 2024 February 2024 atmospheric river, day 2 | 59 mm | 6% | 19% | 76% | 158 |
| August 20, 2023 Tropical Storm Hilary | 59 mm | 11% | 92% | 112% | 137 |
| February 13, 2025 | 56 mm | 7% | 83% | 100% | 171 |
| October 14, 2025 | 55 mm | 37% | 106% | 104% | 182 |
| Rain: ERA5 daily total at the San Fernando Valley point. Storms that lasted several days show low output on the following days too. | |||||
On the wettest days, output fell to a few percent to a third of a clear day. What follows depends on the storm, not the panels: when the sky cleared, the next day was a normal day — after Tropical Storm Hilary in August 2023, roofs were back to 92% the next day and 112% the day after. Multi-day storms such as the February 2024 atmospheric river kept output low until the rain moved on. Whether storms damaged anything is checked in the next section.
Did the storms damage anything?
Each system’s clear-day output for the sunlight, 3–14 days before and after each storm.
| Storm | Systems | After vs before, median system | Systems more than 10% lower after |
|---|---|---|---|
| September 9, 2022 Tropical Storm Kay remnants | 53 | +14% | 2% |
| August 20, 2023 Tropical Storm Hilary | 123 | +8% | 1% |
| October 14, 2025 | 161 | +5% | 3% |
| December 24, 2025 | 172 | +6% | 0% |
| September 6, 2026 Remnants of Hurricane Marie | 177 | +14% | 0% |
| Calm, dry periods (comparison) | 8,460 windows | −2% | 3% |
| Clear days only (ERA5), output divided by the day’s sunlight. Winter storms followed by more cloudy weeks have too few clear days on one side to check and are left out. Comparison: the same windows around dates with no rain within two weeks. | |||
Across 698 system-storms, only 1% came out of a storm more than 10% lower on clear days, fewer than the 3% that show such a drop over any two calm, dry weeks. The typical system came out of a storm 8% better, because the rain had washed the panels (dirty panels, measured). In these storms, wind and rain did not leave a mark on output. Hail, falling branches and roof leaks are rarer events that a fleet median would not show; for those, see how durable solar panels are in storms, and if your own system stays low after the sky clears, check the daily graph and the daily fleet index.
What clouds cost over a year
Adding up every day, a typical roof made about 86% of what it would in a year of only clear days (middle half of system-years 83–88%). In other words, clouds and rain take roughly 14% of the potential, most of it in winter and spring. Solar production estimates for Los Angeles already include this: they are built on typical weather years, not on clear skies. For the yearly totals themselves, see the production benchmark and how much one panel produces per day.
Methodology
How the numbers were produced.
- Data source
- Daily AC energy from Enphase microinverter systems in Cali Energy’s monitoring fleet. Monitoring readings, not utility meter readings.
- Systems and days
- 219 rooftop systems and 241,662 valid system-days, 2019–2026. Days without valid data were removed (zero days, which usually mean no data, the days around each gap, back-filled runs). 31 systems outside our four weather areas were left out.
- Clear-day output of each roof
- For each system and each date, the clear-day level is that roof’s own output on its best days within ten days of the same calendar date (90th percentile of output relative to clear-sky sunlight, all years together), scaled to that day’s clear-sky sunlight. Roof direction, tilt and shade are therefore built in. Results are shown against the median clear day, set to 100%.
- Sky condition
- ERA5 reanalysis (ECMWF / Copernicus, grid of about 25 km) daily sunlight at the nearest of four points: San Fernando Valley, Los Angeles basin, Santa Clarita Valley and Ventura County. Each day’s sunlight is divided by that point’s clear-day level for the same time of year (the 95th percentile of days within ten calendar days, 2001–2026). Rainfall: ERA5 daily totals at the same point.
- Sensitivity
- Sorting days by NASA POWER satellite data instead — a coarser grid of about 100 km that puts all four areas in one cell — gives 62% for a partly cloudy day, 36% for a mostly cloudy day and 16% for dark overcast. The exact percentages depend on how cloudiness is measured; the ranking and the size of the yearly loss do not.
- Limits
- A 25 km weather cell can miss a patch of morning fog or marine layer over one neighborhood, so some days sit in the wrong bin. Systems are those we monitor, not a random sample.
- Updates
- Recalculated with each data update.
Sources: ERA5 reanalysis, ECMWF / Copernicus, via Open-Meteo (open-meteo.com) — daily sunlight and rainfall; NASA POWER daily clear-sky sunlight (power.larc.nasa.gov) — seasonal clear-sky shape and the sensitivity check.
Download the data
https://cali-energy.com/research/solar-output-cloudy-rainy-days-los-angeles. Aggregated data are free to use under CC BY 4.0 with a link to this page.Frequently asked questions
How much do solar panels produce on a cloudy day?
In our Los Angeles measurements, about 70% of a clear day on a partly cloudy day, about 44% on a mostly cloudy day, and about 23% under dark, heavy overcast. Thin or passing clouds cost little: a mostly sunny day gave about 90%.
Do solar panels work on rainy days?
Yes, at reduced output. Days with light rain gave about 63% of clear-day output, moderate rain about 46%, and heavy rain about 23%.
How much do clouds affect solar panels over a year in Los Angeles?
A typical roof made about 86% of what it would in a year of only clear days, so clouds and rain cost roughly 14%, mostly in winter and spring.
How many cloudy days does Los Angeles get?
At the San Fernando Valley point in 2023–2025, about 194 days a year were clear, 100 mostly sunny, 41 partly cloudy and 31 mostly cloudy or overcast, by ERA5 weather data.
Does solar production recover after a storm?
Yes. In our data, output returned to normal as soon as the sky cleared, usually within one to two days. Multi-day storms simply lasted longer.
More research
Next steps
Output low even on clear days?
Prepared by Cali Energy, October 15, 2026. Figures describe systems in our monitoring fleet and are not a guarantee of what any specific roof will produce; orientation, tilt, shade, equipment and weather change results. Not engineering, financial or utility advice. See our Content Disclaimer.