Growing Through Deep Winter: A Comparison of Greenhouse Approaches

Growing Spaces has built geodesic dome greenhouses for more than 35 years. Since buying the company in April 2025, I have been working through a question our customers face: can I grow food through winter where I live, and what kind of greenhouse makes sense?

I have a commercial interest in that question because we sell one of the options. But I can also explain where I think a dome fits and where I would recommend another approach.

What 'Deep Winter Greenhouse' Means

In the Upper Midwest, a deep winter greenhouse usually refers to a specific kind of passive solar greenhouse developed for northern winter production: an east-west structure with most of its glazing facing south, heavy insulation on the north side, and thermal mass to hold daytime solar heat after sunset.

The modern use of the term is closely tied to Minnesota growers and researchers. Carol Ford and Chuck Waibel popularized the approach through The Northlands Winter Greenhouse Manual, and the University of Minnesota refined the concept through its Deep Winter Greenhouse research and design program. The goal was practical: grow cold-hardy crops through a northern winter with minimal added heat and, in many cases, no grow lights.

To make use of low winter sun, a typical deep winter greenhouse combines steep south-facing glazing with an insulated north wall and roof. Many designs move warm daytime air into an underground rock bed, then draw stored heat back out to moderate nighttime temperatures. Some sites are also partially set into grade or earth-sheltered.

That approach is suited to the cold-hardy crops at the center of the University of Minnesota’s program, including greens, herbs, brassicas, and sprouts. Even with that protection, winter light limits growth, especially in December and January. Growers also need to manage ventilation and summer heat as the seasons change.

A Growing Dome approaches the same heat-storage problem with a curved glazed shell, a reflective insulated north side, and an above-ground pond. Its interior has room for different bed layouts and taller plants, although how much of that space can be used through winter still depends on the crops, light, and supplemental heat.

High-Quality Insulation: Comparing the Two Enclosures

In a winter greenhouse, insulation slows the escape of heat collected during the day. By the early hours of the morning, the useful question is how much warmth remains after losses through the glazing, frame, foundation, vents, and air leaks. Both the Minnesota deep winter greenhouse and the Growing Dome combine insulation with thermal mass, but they make different choices about how much of the enclosure admits sunlight.

A glazing comparison helps put that choice in perspective. For 100 square feet of glazing with a 30°F temperature difference between inside and outside, conductive heat loss is approximately:

• Single-pane glass at about R-0.9 loses roughly 3,300 BTU per hour
• Twin-wall polycarbonate at about R-1.7 loses about 1,800 BTU per hour
• 16 mm multi-wall polycarbonate at about R-2.8 loses about 1,100 BTU per hour

At these illustrative R-values, moving from single-pane glass to 16 mm multi-wall polycarbonate cuts conductive heat loss through the glazing by roughly two-thirds. That is a substantial improvement, but glazing still loses more heat per square foot than a heavily insulated opaque wall. These figures compare the materials at the same area and temperature difference; a whole-greenhouse comparison also has to account for how much of each material is used and how much air leaks through the enclosure.

The University of Minnesota designs put much of their insulation where little useful winter sunlight would enter, while concentrating glazing on the south face. In the v2.0 structures described in a study of the 2018 growing season, the researchers specified R-23 above-grade walls, an R-30 roof, and 16 mm triple-wall polycarbonate set at 60 degrees on the south side. A fan moved heated air into an insulated underground rock bed for storage. That combination gives the grower a heavily insulated enclosure around a concentrated area of winter glazing.

That study, Yield of Leafy Greens and Microclimate in Deep Winter Greenhouse Production in Minnesota, monitored three greenhouses during winter and early spring 2018. The reported average air temperatures were about 52°F in Milan, 61°F in Finland, and 63°F in Madelia, with readings across the sites ranging from roughly 36°F to 122°F. The authors attributed the highest temperature to improper ventilation during the lead farmer’s absence. Their system description also included backup heat set to about 45°F. These are useful field results, with differences in design and management to consider when comparing the sites.

The Growing Dome uses 16 mm multi-wall polycarbonate across its curved shell, with reflective insulation lining the north side. The north lining reduces heat loss through the covered area and redirects some sunlight toward the growing space, pond, and soil. Its above-ground pond provides accessible thermal mass, while the glazed portions of the shell admit light from several directions. Compared with the Minnesota approach, more of the enclosure serves as a light-collecting surface instead of a heavily insulated opaque wall.

Under favorable winter conditions, we have observed well-prepared Domes holding about 20 to 30°F above outdoor temperature, depending on available sun, stored heat, and the condition of the enclosure. At 0°F outside, that difference would put the interior around 20 to 30°F; at -20°F outside, it would leave the interior around 0 to 10°F. Those are very different conditions for a plant, even though the temperature difference is the same. I would choose crops and supplemental protection around the indoor minimum they can tolerate, allowing for cloudy periods when less solar heat is available to store.

The Minnesota study reports measured indoor conditions at particular sites, while our 20 to 30°F observation describes a difference from outdoor temperature under favorable conditions. I would not rank the two greenhouses by putting those numbers next to each other. A direct performance comparison would need the same weather, indoor target, operating practices, and accounting for supplemental heat.

For someone focused on cold-hardy winter production, I would take the Minnesota design’s insulated walls and roof seriously. For someone planning a greenhouse garden with light entering from several directions, the Dome offers a different enclosure and layout, with heating needs to assess accordingly. In either case, a thicker panel can only do so much if the site is shaded or the vents leak. I would compare winter sun, enclosure heat loss, air sealing, and thermal mass together before deciding how much supplemental heat the project needs.

Solar Heat Gain: South-Facing Glazing vs. a Dome

Once we know where heat can escape, the other half of the comparison is how sunlight enters. The angle of the glazing affects direct solar collection, while its orientation determines which parts of the day it faces the sun. That makes a steep south-facing wall and a curved dome meaningfully different even when both use multi-wall polycarbonate.

The Minnesota v2.0 design described in the study uses a long east-west orientation with its main glazing set at 60 degrees toward the south. That arrangement is directed toward low winter sun, particularly around solar noon. Earlier and later in the day, the sun reaches the same wall from a more oblique direction. For a grower whose priority is winter greens, concentrating glazing on that productive south face allows the rest of the enclosure to remain heavily insulated.

A dome presents east-facing, south-facing, and west-facing panels to the sun over the course of the day, so different parts of the shell are favorably oriented at different times. This spreads the opportunities for direct collection across the curved surface. The diffused multi-wall polycarbonate then scatters some of the light transmitted through those panels. Geometry affects how sunlight reaches the enclosure; diffusion affects how the transmitted light is distributed inside.

That distinction also matters when comparing glazing. Our published panel specifications list an R-value of about 2.8 and 65 percent light transmission for the 16 mm five-wall material, compared with about R-1.7 and 81 percent transmission for the listed 8 mm twin-wall material. The more insulating panel admits a smaller share of the incoming light in that comparison. Diffusion can spread transmitted light through the growing space, but it cannot replace the light that never passes through the panel.

In our earlier clear-sky geometry model at 40°N, we compared a curved dome surface with a single south-facing plane set at 60 degrees across the late-October to February period. That model estimated about 15 to 20 percent more daily solar exposure on the curved surface. I would keep that result within its scope: it was a geometry estimate, rather than a measurement of crop growth or heating-energy savings. The area being compared, glazing properties, and shading all matter when applying a surface-exposure calculation to an actual greenhouse.

For heating, I would go a step further and compare the solar heat collected with the heat lost through the enclosure over the same period. An additional glazed surface can admit sunlight during the day and lose heat after sunset. The Minnesota design limits much of that exposed glazing to its south face; the Dome pairs its broader glazed shell with north-side insulation and thermal mass. Neither the peak sun angle nor the daily exposure estimate alone settles which structure will need less supplemental heat.

The site can also change which hours of collection are useful. Trees or a mountain ridge may block morning sun, while a nearby building can shade the greenhouse later in the day. Inside either structure, tall crops and shelving affect the light reaching lower plants. In the Minnesota trial, the lower shelves received less light and the upper shelves generally produced higher yields. That is a reason to consider the planting layout along with the shell geometry.

This is where I would bring the comparison back to the grower’s plans. A Minnesota deep winter greenhouse concentrates winter collection through one main glazed face and limits heat loss through the rest of the enclosure. That is a compelling arrangement for cold-hardy winter production. A Growing Dome admits light through panels facing several directions and offers a different interior layout, which may better suit someone planning a varied greenhouse garden through the year. I would weigh those uses alongside the site’s winter sun and heating requirement, rather than choose from shape alone.

Efficient Thermal Mass for a Deep Winter Greenhouse

Once sunlight has warmed the greenhouse, it helps to have somewhere to store that energy. Thermal mass absorbs heat as the greenhouse warms and releases some of it as conditions cool, with the amount available overnight depending on how much was stored during the day.

Water is useful for this because it can hold a lot of heat without changing temperature quickly. A gallon weighs about 8.34 pounds, and each pound stores 1 BTU per °F change, giving us a simple way to estimate the energy involved:

Gallons of water × 8.34 × temperature change in °F = BTUs stored or released

For example, 1,250 gallons of water stores or releases about 10,400 BTUs for every 1°F change in temperature. A 5°F change would therefore represent roughly 52,000 BTUs, while a 10°F change would represent about 104,000 BTUs. The calculation tells us the energy associated with those temperature changes; how much a pond actually warms on a particular day is a separate question.

The table applies the same calculation to several water volumes. For a specific pond, use its actual fill volume; the 22-foot example uses about 900 gallons in a nominal 1,000-gallon pond.

Actual heat exchange depends on sunlight, water temperature, air temperature, wind, snow cover, vent sealing, plant canopy, and how much pond surface area is exposed. A covered pond, a pond shaded by tall plants, or a stretch of cloudy days will behave differently from an open dark pond on a clear winter day.

In a Growing Dome, the pond sits along the north side beneath the reflective insulation, leaving the center of the greenhouse open. Although it also adds humidity and can support aquatic plants or fish, its contribution to winter heating comes from the solar heat it absorbs and can later release.

Snow Load Capacity: What the Rating Depends On

Heat retention is only part of choosing a winter greenhouse; the structure also needs to handle the snow at its site. For a dome, snow-load capacity depends on size, foundation, connections, and any engineered modifications, and its rating needs to be compared on the same design basis as the site requirement.

Our published standard snow values range from 30 psf for the 42-foot Dome to 85 psf for the 15-foot Dome, with intermediate values for the other sizes. These are Flat Roof Design Values (Pf), and higher values require engineering modifications. The current weather-resistance specifications set out the values and design assumptions by size.

The difference between sizes reflects the spans involved: smaller domes have shorter spans and generally carry higher snow loads within this material system, while larger domes cover more area. Foundation and connection details also affect how those loads reach the ground.

For a snowy site, the local requirements need to be checked against the applicable drawings, particularly because ground snow load and roof design load are different quantities. Depending on that review, the project may call for a particular foundation, reinforcement, or stamped plan set.

The curved roof can shed snow, but accumulation and drifting still need to be accounted for in the engineering and maintenance plan.

Supplemental Heating: What It Can Cost Through Winter

Once the structure is suited to the site, people want to know how much heat they will need after sunset and what it will cost. I would start with the minimum temperature their crops need, then look at local winter weather and the greenhouse’s heat loss.

The Growing Dome uses solar gain, insulation, thermal mass, and ventilation to reduce reliance on mechanical equipment, but the indoor temperature target still has a large effect on heating needs. Holding a minimum in the mid-50s requires less heat than holding 70°F through the same winter, and a lower target for cold-hardy crops changes the calculation again.

Air circulation helps distribute the available warmth. The undersoil system moves air from near the pond through ducts under the beds and out toward the south side, helping distribute pond-moderated air and reduce temperature stratification. When planning supplemental heat, count it as an air-circulation system.

To illustrate the heat-loss calculation, the original version of this article used an earlier model with an effective U-value of 0.56 BTU per hour per square foot per °F and a 1,119-square-foot shell area for a 26-foot Dome. I have retained that historical example below, although our current HVAC methodology uses revised geometry and heat-loss assumptions for sizing work.

Earlier illustrative heating load = 0.56 × 1,119 × indoor-outdoor temperature difference

Using those earlier inputs to hold 55°F, before accounting separately for solar gain, stored heat, and air leakage:

To turn that heating load into electricity use, we also need to account for equipment performance. At a coefficient of performance (COP) of 3, a heat pump delivers three units of heat per unit of electricity, compared with about one for resistance heat. Because heat-pump performance changes with conditions, a cold-weather equipment selection needs more than a seasonal average.

The seasonal cost estimates below accompanied that earlier model, assuming a 26-foot Dome with a pond, a heating target in the mid-50s, and a heat pump supplying most supplemental heat from November through March. The cost calculation used electricity rates of 16.79¢/kWh in Colorado, 15.39¢/kWh in Minnesota, and 25.79¢/kWh in Alaska. These historical examples have not been recalculated under the current methodology.

Those examples show how operating costs change with climate, electricity price, and growing temperature. For a current project budget, local rates and the updated sizing assumptions are the starting point, with site conditions considered alongside them. Shading, snow-covered glazing, air leakage, and a higher temperature target can all increase the heating requirement.

It is also important to size the heater for times when stored solar heat is unavailable. In a cold, dark winter, mechanical heat may run regularly even in a greenhouse designed to use passive systems first.

Adaptability to Local Climate: Field Notes From Real Domes

The field accounts below show what those decisions look like in use, as owners adapt to winter light, humidity, wind, snow, and their growing goals. Their choices can matter as much as the headline outdoor temperature.

That is why I don’t treat climate adaptation as a single upgrade. Depending on the site, the useful change might be more ventilation, additional thermal mass, a different foundation, backup heat, grow lights, shade cloth, misting, or a different crop plan. Sometimes owners remove or replace a standard component when local conditions call for another approach, as several of these installations illustrate.

Haines, Alaska: heavy wet snow and humid summers

Mardell Gunn and Mark Kistler chose a 26' Growing Dome after years of gardening with hoop houses in Haines, Alaska. Their site needed a structure that could handle heavy snow, wind, and humid summers while creating enough protection for heat-loving crops. In one winter, they reported nearly six feet of wet, heavy snow on the greenhouse at a time. They added a center support pole, extended the entry to relieve snow pressure near the door, skipped the standard pond to reduce humidity, and built black concrete raised beds as their thermal mass. They added extra ventilation and used the undersoil fans to reduce mold pressure. Because winter light is limited and electricity is expensive, they don’t treat the dome as a true four-season production greenhouse. They use it aggressively from late May through October, producing 150 to 200 pounds of tomatoes along with peppers, cucumbers, and tall squash.

Wrangell, Alaska: maritime humidity and low winter light

Laura and Dwane’s 22-foot Growing Dome sits near the Pacific Ocean on Wrangell Island. Their account describes humid conditions and limited winter light. Tomatoes, eggplants, jalapeños, potatoes, and Swiss chard performed well; lettuce struggled, and mushrooms appeared inside. They reported an interior around 50°F through the year, but limited sunlight from October through February constrained winter production. The account shows why temperature alone cannot describe a greenhouse’s growing conditions.

Naujaat, Nunavut: food production at the Arctic Circle

Green Iglu, formerly Growing North, began its first project in Naujaat, Nunavut, in 2015. The pilot used two 42' Growing Domes for a remote community of about 1,200 people located directly on the Arctic Circle. This is a different use case from a backyard winter greenhouse. The goal is local food access, training, and community-scale production in a place where fresh produce is expensive and supply chains are fragile. After the pilot, the greenhouses were taken over by local Arctic farming communities and continued producing year after year. Later Green Iglu projects expanded into other remote Canadian communities including Lax Kw'alaams, British Columbia; Arviat, Nunavut; Port Hope Simpson, Newfoundland and Labrador; and communities in Quebec.

Steuben, Maine: cold, snow, and shoulder-season harvests

Velma Orcutt’s Maine Growing Dome has operated through heavy snow and winter temperatures around -15 to -10°F. She has used it to start seedlings, harvest vegetables after fall frost, and grow peppers, tomatoes, lettuce, cucumbers, and other garden crops with more protection than outdoor beds provide.

Lacombe, Alberta: school-scale production and education

Lacombe Composite High School installed a 33' Growing Dome as a student-led project after the school's original solar project was lost in a roof fire. The dome became part greenhouse, part classroom, and part sustainability lab. With outside greenhouse expertise, the students added a climate battery and grew tropical and warm-season crops including pineapples, sweet potatoes, bananas, figs, and lemons. Later, the school added an aquaponics project using the above-ground pond and reused tank water to reduce water use in the growing cycle.

The Growing Dome vs. Other Deep Winter Greenhouses

Looking across those examples, I come back to what you want it to do. There is no single best winter greenhouse for all of those uses.

If the aim is winter greens with limited supplemental heat, I would look closely at the Minnesota deep winter greenhouse designs, whose steep south glazing, insulated enclosure, and underground heat storage are directed toward that purpose.

If the priority is a lower initial investment in spring and fall growing, I would also consider a hoop house. In that case, I would make sure the budget and maintenance plan account for the film, frame, anchoring, and expected weather exposure.

The Growing Dome fits someone who wants a durable greenhouse garden with room for different crops and uses through the year. If those plans include warm-season or tropical crops in winter, though, they may require substantial heat and supplemental light, so the equipment needs to be chosen with that use in mind.

If a Growing Dome is closer to what you need, our four-season greenhouse kit guide helps you choose a size and match the configuration to your crops, climate, and plans for heat or light.

When comparing costs, I would work from completed-project budgets. A Growing Dome kit price leaves out site work and other construction costs that may be included in a deep winter greenhouse study, so similar headline prices can conceal a difference in what the project includes.

A quick reference summary:

Note: University of Minnesota figures drawn from University of Minnesota Extension's 2018 "Winter greenhouse enterprise analysis," based on 8 Upper Midwest deep winter greenhouse enterprises. Original 2018 figures were $18,500 average construction cost and $33/sq. ft.; both have been adjusted to early-2026 dollars using CPI (~33% cumulative). The sample is small and not statistically significant; treat as a study finding rather than a current market estimate. Construction-specific inflation has likely run higher than general CPI over this period, so these adjusted figures may still understate current build costs.

If you want to talk through a site or growing plan, send us a note with your crop goals and what you know about the local conditions. That gives us a useful starting point.

From there, the beds, paths, work area, and equipment help determine how much space will be useful. A smaller dome may fit the plan well, while a larger one may make the interior easier to use, provided it can meet the site’s snow-load requirements.

For sites with unusual load requirements, we offer paid engineering customization. That review may lead to reinforcement, a different foundation, or a smaller dome.

-Gary