How Much Does Solar Production Drop in Winter? Measured in Southern California, Rainy and Dry Winters
On the same rooftop system, December produces a median 46% of June across 768 system-years on two monitoring platforms; the middle half ran from 38% to 55%. November through February together make about 23% of a year’s solar. Truly dark days are few: a median 3 to 9 per winter, depending on the year. And a rainy winter did not mean a weaker one: the wet winter of 2022–23 and the dry winter of 2024–25 produced almost the same, though the dry one had 12% more sun.
How deep the winter drop is on each system
December as a percent of the same system’s June.
| Measure | p10 | p25 | Median | p75 | p90 |
|---|---|---|---|---|---|
| December as % of June | 32% | 38% | 46% | 55% | 62% |
| December as % of July | 31% | 36% | 43% | 50% | 55% |
| Nov–Feb share of the year | 20% | 21% | 23% | 25% | 26% |
| Best month ÷ worst month | 1.9× | 2.1× | 2.4× | 2.9× | 3.6× |
| Both platforms, 2019–2025: SolarEdge 381 and Enphase 387 system-years. December as % of June: SolarEdge 45%, Enphase 46%. | |||||
https://cali-energy.com/research/winter-solar-production-southern-california. Aggregated data are free to use under CC BY 4.0 with a link to this page.The spread is the finding. On a typical system December makes a bit under half of June. On one in ten it makes less than 32%, and on the best tenth more than 62%. The difference between those systems is not the weather — they share it — but the roof: which way it faces, how steep it is, and what stands between the panels and a low winter sun.
Rainy winters, dry winters
November–March. Output: median kWh per kW of systems measured in every winter shown.
| Winter | Rain, Oct–Apr (ERA5, Valley) | Sunlight, Nov–Mar (ERA5, kWh/m²) | Dark days, median system | Output, Enphase (47 systems) | Output, SolarEdge (31 systems) |
|---|---|---|---|---|---|
| 2020–21 | 122 mm | 622 | — | — | 530 |
| 2021–22 | 316 mm | 611 | — | — | 503 |
| 2022–23 | 758 mm | 525 | 9 | 448 | 438 |
| 2023–24 | 555 mm | 549 | 6 | 442 | 437 |
| 2024–25 | 198 mm | 586 | 3 | 437 | 429 |
| 2025–26 | 536 mm | 578 | 8 | 448 | 442 |
| Dark day: under 20% of the same system’s clear-day output for that time of year (Enphase daily data, 2022 onward). Rain and sunlight: ERA5 at our San Fernando Valley weather point. Output: kWh per kW, November–March, systems with all winters in the column. | |||||
Rain sets how many dark days a winter has: 9 for the median system in 2022–23, when our Valley weather point recorded 758 mm, against 3 in 2024–25, with 198 mm. It did not set the winter’s output. The dry winter had 12% more sunlight, yet the same Enphase systems made 2% less (437 against 448 kWh per kW), and the SolarEdge systems tell the same story. Part of the answer is dust: rain also washes panels, and after the long dry spell of 2024 the first soaking rain in late January 2025 lifted the same roofs to 104–109% of their December output for the sunlight (smoke, ash and the first rain; dirty panels, measured). The rest belongs to a wider shortfall since 2024 that the sunlight data do not explain, followed in why solar production drops from year to year. The 2023–24 winter, an El Niño year, was not different from the others.
Real roofs lose more in winter than the models say
Measured median as a percent of the PVWatts ideal-roof estimate.
| Month or period | Measured median, kWh/kW | PVWatts ideal roof | Measured as % of ideal |
|---|---|---|---|
| June | 145 | 164 | 88% |
| July | 158 | 172 | 92% |
| December | 69 | 100 | 69% |
| January | 80 | 110 | 72% |
| Best-to-worst month | 2.3× | 1.7× | deeper swing on real roofs |
| All twelve months: see the monthly benchmark. | |||
Three things make winter harder on real roofs than in the model. Shade: in December the sun stays low in the southern sky, so trees, a neighbor’s second story or a chimney can shade panels that are clear all summer; PVWatts assumes no shade. Direction: east- and west-facing arrays lose a larger part of the short winter day than south-facing ones. Weather and dust: storms and dusty dry spells come in some years and not others, while the model uses a typical year and clean panels. The pattern — close to the model in summer, well below it in winter — is consistent across the fleet.
Where does your winter drop sit?
Place your December-to-June ratio among the measured systems.
| December as % of June | Where it sits among measured systems | Common reasons on real roofs |
|---|---|---|
| Above 55% | Above the middle half of observed seasonal ratios | South-facing panels with little winter shade |
| 38–55% | Within the middle half of observed seasonal ratios | Typical mix of roof direction and seasonal shade |
| 32–38% | In the lowest quarter | East- or west-facing panels, winter shade from trees or buildings |
| Below 32% | In the lowest tenth | Heavy winter shade, or something that changed after June |
| A seasonal ratio is a screening signal, not a diagnosis of the equipment. | ||
Suggested next step. Use the same system’s June and December of the same year. In the lowest quarter, compare with last winter and look at what may have grown or been built nearby. In the lowest tenth, also check the daily graph in your app for flat or zero days and any inverter alerts, and consider a check. A ratio that falls from one winter to the next while the weather was similar points to something new: a tree, a dirty or damaged panel, or equipment. For a single month against all LA systems, use the monthly check; for a single dark day, the daily fleet index.
What the winter dip means for your bill
Winter bills come back
Lower solar output plus heating and holiday use means more grid power in winter. See solar panels in winter.
Credits are banked in summer
Under net metering, summer credits offset winter use at true-up. See the NEM true-up bill.
Size batteries for winter too
A battery that fills every summer day may not fill in December; see how full LA batteries get.
Shade is fixable
Trimming a tree can restore winter output; see solar, shade and trees.
Methodology
- Systems
- 264 systems on Enphase and SolarEdge monitoring with complete calendar years 2019–2025, under the rules of the platform comparison: 12 valid months, no downtime year, a plausible output per kW and a verified DC rating.
- Seasonal ratios
- Within each system-year: December ÷ June of the same system and year. 4 system-years where December exceeded June or July, a sign of a summer outage, were excluded.
- Dark days
- Enphase daily production checked day by day. A system’s clear-day level for each date is the 90th percentile of its output within 10 days of that date across all years; a dark day is under 20% of it. Counted November–March for systems with at least 140 valid days that winter.
- Winters compared
- November–March kWh per kW for the same systems in every winter of the column (47 Enphase, 31 SolarEdge). Rain and sunlight from ERA5 at our San Fernando Valley weather point.
- Limits
- Not a random sample of LA roofs. Roof direction, tilt and shade are not adjusted for.
Frequently asked questions
Do solar panels work in winter in Southern California?
Yes, at reduced output. In our measured fleet December produced a median 69 kWh per kW in 2023–2025, about 46% of the same system’s June.
How much less do solar panels produce in winter?
On the same system, December makes a median 46% of June (middle half 38–55%). November through February together make about 23% of the year’s output.
Does a rainy winter mean less solar?
Not in our data. The rainy winter of 2022–23 had more dark days (9 against 3) and less sun than the dry 2024–25, but the same systems made about the same: 448 and 437 kWh per kW. Rain also cleans the panels.
Why is my solar so low in December?
Short days and a low sun. Shade from trees and nearby buildings reaches much farther in December, and east- or west-facing panels lose more of the day.
What is a normal winter drop for solar in LA?
Among the systems we measured, the middle half made 38–55% of their June output in December. Below 32% puts a system in the lowest tenth — a screening signal worth a closer look, not a diagnosis.
Is solar production lower than PVWatts estimates in winter?
In our data, yes: about 69% of the ideal-roof PVWatts figure in December versus 92% in early summer, because the model assumes no shade and clean panels.
More research
Next steps
Planning for winter, or your December looks too low?
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.