Weather-Adjusted Solar Production: Why "Last Year" Is a Bad Baseline

By Olivier Beauchemin · Updated May 2026

Your solar production was lower this year than last year. Is your system broken, or was it just a cloudier year? Most homeowners can't answer that question — and that uncertainty is exactly why underperformance goes undetected for years.

The problem is the baseline. To know whether your system is performing correctly, you have to compare its output to something. The two baselines most people reach for — last year, and the installer's original estimate — are both unreliable for the same reason: they don't account for the weather that actually happened.

Why Year-Over-Year Comparison Fails

Sunlight is not a constant. The amount of solar energy reaching your roof — what the industry calls irradiance — varies meaningfully from one year to the next. A cloudier-than-average year, a wetter spring, more wildfire smoke, or a snowier winter all reduce the energy available to your panels.

Year-to-year swings of several percent in available sunlight are normal and have nothing to do with your equipment. So if your production drops 6% from last year, that could be:

• A genuinely cloudier year, with your system working perfectly
• A perfectly normal year, with a real 6% problem in your system
• Some combination of the two

A raw year-over-year number cannot tell these apart. It mixes weather and system performance into a single figure and gives you no way to separate them. A homeowner who sees a small decline and shrugs — "must have been a cloudy year" — may be right, or may be ignoring a failed component.

Why the Installer's Estimate Fails Too

Your installer's original production estimate — the number on your proposal — is built from average weather. It typically uses a "typical meteorological year," a synthetic dataset that blends decades of weather into one representative year.

That's a reasonable way to estimate long-term output before installation. But it's a poor way to judge a single real year. The actual year you just lived through was not the average year. If it was sunnier than average, your system can underperform and still beat the estimate. If it was cloudier than average, your system can be perfectly healthy and still fall short of the estimate.

Comparing one real year against an average-weather estimate guarantees noise. You'll get false alarms in cloudy years and false reassurance in sunny ones.

What "Weather-Adjusted" Actually Means

A weather-adjusted baseline answers a more precise question: given the sunlight that actually reached your specific roof this period, how much energy should a healthy system of your size and orientation have produced?

Building that number requires:

• Actual local irradiance data for the period in question — not an average year, but the real weather, sourced from satellite-derived or ground-station measurements for your location
• Your system's specifications — capacity in kilowatts, panel tilt, azimuth (the compass direction your roof faces), and known shading
• A physics-based model that converts available sunlight into expected AC energy, accounting for temperature effects, inverter efficiency, and standard system losses

The result is an apples-to-apples comparison. When your weather-adjusted expected production says 1,150 kWh for the month and your inverter reported 1,000 kWh, the 13% gap is not weather — the model already removed weather from the equation. That gap is your system.

The Same Logic Applies to Production Guarantees

If your contract includes a production guarantee, the weather-adjusted question becomes a financial one. When you tell your installer "my system fell short this year," the first thing they will say is "it was a cloudy year." Without a weather-adjusted baseline, that ends the conversation. With one, you can show that the shortfall remains after weather is accounted for — which is the difference between a defensible claim and a dismissed one.

What to Look For

You don't need to build a physics model yourself. But you should understand what a credible weather-adjusted comparison includes, so you can tell whether a tool or report is giving you a real answer:

• It uses actual measured weather for the period, not a long-term average
• It is calibrated to your specific system — size, tilt, orientation — not a generic estimate
• It reports the gap in kilowatt-hours and dollars, so you know whether it's worth acting on
• It can be applied to short periods — a month or a quarter — so you catch problems early, not at a year-end reconciliation

How OwlWatt Does This

OwlWatt connects to your existing inverter monitoring and continuously compares your actual production to a weather-adjusted expected baseline. The baseline uses NREL-validated irradiance modeling and the real weather for your location, calibrated to your system's capacity, tilt, and orientation.

When your production falls below what the weather can explain, you get an alert with the estimated dollar impact — not a vague "lower than last year," but a specific, weather-adjusted answer to the only question that matters: is your system actually working?