Depth is the wrong measurement
The question people ask is how much snow a roof can take, and they ask it in inches. Inches are almost useless on their own. Freshly fallen cold snow is mostly air, and a foot of it can hold less than an inch of water. The same foot after a week of settling, a thaw and a refreeze can hold three inches. The weight difference between those two states is a factor of three, and nothing about the depth on the ruler changed.
What does the work is the water equivalent: the depth of liquid water you would be left with if the whole pack melted in place. That figure converts directly, with no assumptions and no fudge factor, because water weighs 62.4 pounds per cubic foot. Divide by twelve and one inch of water spread over one square foot weighs 5.2 pounds. Six inches of water equivalent is 31 pounds per square foot, whether it arrived as four feet of powder or eight inches of slush.
The bands in the selector above are wide on purpose. Anyone quoting a single density for snow is quoting the middle of a range that spans an order of magnitude, and the range is the honest answer.
Measuring it yourself instead of guessing
The guess can be replaced with a measurement in about twenty minutes, and it needs nothing but a straight-sided container. Push the container down through the full depth of the pack somewhere flat, cut it free at the base, bring it inside and let it melt. Measure the water depth, divide by the snow depth you sampled, and you have your own water content percentage. Enter it in the custom field and the range collapses to a number.
Two things make the sample worth taking. The first is that regional rules of thumb are wrong more often than they are right, because the same storm deposits different snow at different elevations and on different days. The second is that a pack that has been through a thaw is not the snow that fell, and the only way to know what it has become is to melt some of it.
The way loads suddenly get worse
| Event | Effect on the load | Why it catches people out |
|---|---|---|
| Rain onto a snowpack | Adds its full depth as water, 5.2 psf per inch | The depth on the roof barely changes, or goes down, while the weight climbs |
| Thaw and refreeze | Same water, far less depth | The pack looks smaller and is exactly as heavy |
| Drifting | Concentrates several times the average in one place | Happens at walls, dormers, parapets and roof steps, which are often the weakest spots |
| Sliding from an upper roof | Dumps one roof onto another | The lower roof was never carrying that area |
| Blocked drains on a low slope | Ponded water at 5.2 psf per inch | Deflection makes the pond deeper, which makes the deflection worse |
The last row deserves its own sentence, because it is the mechanism behind most low-slope failures. Water collects at the low spot, the added weight deflects the deck, the deflection deepens the low spot, and more water collects. It is a loop that does not settle.
What this page will not do
It will not tell you a roof is safe, and you should mistrust anything that does. This calculator produces one side of a comparison — the demand, in pounds per square foot. The other side is capacity, and capacity is a property of a specific structure: member sizes, species and grade, actual clear span, spacing, connections, condition, whether the framing is rafters or engineered trusses, and what has been added to it since it was built. Design snow loads are set by the code adopted in your jurisdiction and vary enormously between locations, so no figure of that kind appears anywhere on this page.
What the number is genuinely good for is the conversation. Walking into a building department or an engineer with a measured water equivalent and a load in pounds per square foot is a completely different conversation from walking in and saying there is a lot of snow up there.
Before anyone climbs
Clearing snow off roofs injures and kills people every winter. The two mechanisms are falls, which need no explanation, and roof rakes touching overhead service conductors, which is the one people do not see coming because they are looking up at the snow rather than out at the wires. This page deliberately contains no removal technique and no ladder advice. If a roof has to be cleared, it is work for someone equipped, trained and insured for it.
Ice at the eave is a separate matter with a separate cause. It forms because heat is escaping into the roof deck, melting the underside of the pack, and the meltwater is refreezing where the deck is cold. The snow is the raw material, not the reason, which is why removing ice from the gutter fixes nothing that lasts. That failure lives in the insulation and ventilation of the space below, and the home insulation guide is where the fix is, with the heat loss calculator to size how much heat is escaping in the first place. When the storm has passed, the snow removal time calculator handles the ground-level half of the job and the ice melt calculator covers the walks.
Questions people ask
How much does snow weigh per square foot?
Between roughly 1 and 5 pounds per square foot per inch of depth, and the spread is the point. Fresh cold powder can be under half a pound per inch of depth per square foot; settled snow runs around 1 to 1.3; wet or rained-on snow and old packed drifts run 1.8 to 2.6; solid ice is about 4.8. The reliable version of the question is not about depth at all. Melt a core sample, measure the water, and use 5.2 pounds per square foot per inch of water. That conversion has no error bars in it, which is why this calculator is built around it.
Should I use the sloped roof area or the footprint?
The footprint — the outline of the roof projected onto the ground. Snow falls vertically and accumulates on a horizontal basis, so a square foot of ground beneath a steep roof catches the same snow as a square foot beneath a flat one, spread over more roof surface. This is the opposite convention from shingles or sheathing, which are quantified on the sloped surface and are the reason the roofing squares calculator asks for pitch. Using sloped area here would overstate the total weight, sometimes badly.
Does a steep roof mean I do not have to worry?
It means less snow tends to stay, which lowers the accumulated load, and it introduces a different problem. Snow released from a steep roof comes off as a mass rather than gradually, and it arrives on whatever is beneath: entries, walkways, decks, meters, vehicles, and occasionally people. Steep roofs also still drift on the lee side and still build load in valleys and against dormers. Slope changes the shape of the risk rather than removing it, and it does not change whether the framing under it is adequate.
What are the signs a roof is in trouble?
This is one of the questions where a web page should hand you off rather than answer. Descriptions circulate of doors that stop closing, new cracks in ceiling finishes, sagging visible along a ridge or ceiling line, and noises from the structure. Any of those warrants getting people out and calling a structural engineer or your building department the same day, not comparing a number on a screen. What this calculator gives you is a measured load figure to bring to that call, which is more useful than a description of the snow.
Why does the calculator give a range rather than a number?
Because the input is a snow type rather than a measurement, and every snow type covers a real spread of water content. Reporting a single figure from a category would be inventing precision that the input does not contain, and the low and high ends of these bands differ enough to matter. If you want one number instead of two, melt a core sample and enter the measured water content in the custom field. Then the range collapses, and it collapses onto something you actually measured rather than something the page assumed.