A Colorful Late-Fall Ice Show on a New Hampshire Pond

Photography by Steven Arcone.

A note from the editors: An abridged version of this photo essay appears in the Autumn 2023 print issue of Northern Woodlands magazine. We were fascinated by Steven Arcone’s detailed study of early season ice formations and hope that you enjoy them, too. Arcone, a retired geophysicist and adjunct professor at Dartmouth College, has made 12 research trips to Antarctica, lost count of those to Alaska, and has published 80 journal articles – mainly on radar exploration of glaciers, glacial sediments, and permafrost.

“Stick season,” the month or so between the mid-autumn foliage color-fest and winter’s first sparkling snows, might seem an unappealing time for outdoor photography. However, by mid-November, I anticipate a colorful treat from a small pond near my home in West Lebanon, New Hampshire. The pond has no formal name, so I call it Chambers Pond because it lies within the 19-acre Chambers Memorial Reserve. It is about 240 yards east of, and about 170 feet above, the Connecticut River and sits on a bedrock plateau covered with glacial debris. On a sunny morning after an early freeze, the pond displays a variety of ice formations, many resembling branches, stars, feathers, and grids. Blue and yellow light scatters from these formations, and still-colorful leaves are visible just beneath the pond’s frozen surface.

Last autumn, I documented three episodes of surface ice formations between November 15 and December 9. I used a variety of camera angles and lenses, including macrophotography of near-shore formations and of more distant areas. I calculated the size of off-shore ice formations based on such considerations as focal point distance and shoreline reference objects, and I directly measured the size of near-shore images.

Ice skims can form on any water body on a cold night. Calm, fresh water does not spontaneously freeze when its temperature drops below 32 degrees, but in this “supercooled” state, it’s subject to rapid freezing as temperatures continue to drop or a light breeze disturbs the surface. “Seed” particles such as dust or larger objects that protrude through the water’s surface create platforms for crystalline growth, a process called nucleation. What makes Chambers Pond or any small, shallow pond special for spectacular mid-autumn ice formations are the many branches, twigs, stems, and leaves that break the surface and provide nucleation sites.

Recrystallization, melting, and snow can all ruin these formations very quickly, so don’t delay looking for them when the temperature drops.

An early autumn 2022 view of Chambers Pond looking almost due south. The maximum width and length are about 75 and 200 feet, respectively. The maximum depth of about 4 feet occurs at the far, southern end. The pond lies in a 10- to 40-foot depression, which helps trap cold air.

When I arrived at Chambers Pond at 8:30 during the sunny morning of November 15, the air temperature was 24 degrees. An ice skim completely covered the pond surface, showing dendritic (branching) starlike and feathery formations. This wide view of the pond looks west-northwest from the southern shore. The thinnest ice was where reflections gave the false impression of open water.

This telephoto shows the starlike formation in the center of the previous wide view and the thin twig from which it first grew. Between the star arm blades are smaller dendritic forms that developed as the main blades grew outward. Blue areas are where red and green wavelengths were absorbed during reflections within the ice before they reached my camera. Yellow areas are where surfaces sufficiently scattered all sunlight to mask any of these returning internal reflections. The horizontal arm is roughly 110 centimeters long. In general, blades can well exceed 1 meter in length.

Taken during the overcast afternoon of November 15, this image from the now colorless ice shows gridded, curvilinear mosaics that were likely seeded by particles in the water. The fernlike patterns across the top grew from the blade that extends across the image, as did the mosaics bordering the long sides of the large triangle. The horizontal midline is about 125 centimeters wide.

Warm temperatures ruined the original crystalline forms by November 17. By November 25, when this image was taken, the southern area retained at least 5 centimeters of wet ice and had developed a different starlike formation comprised of clear black ice, each surrounded by a polygon of cloudy, blue-tinted ice.

By November 30, bubbly, wet ice remained in the southern area, a black ice skim remained in the northern area, and there was no ice in the middle of the pond. On December 1, the early morning temperature dropped to 33 degrees; radiational cooling (emission of infrared “heat rays” from the Earth’s surface into the atmosphere) and gusty overnight winds surely caused further lowering of the water temperature. By morning, when this image was taken, a sea of approximately 1- to 15-centimeter-long blades had nucleated from the abundant leaves, stems, and other material, and now covered the middle of the pond, but not quite to the shores.

This nearly overhead detail, from December 1, is from the near edge of the mid-pond sea of blades. The interior of some of the polygons were perhaps unfrozen because they show no associated dendritic mosaics.

Many transparent stars occurred along the north shore during the first five days in December. The central brown twig within the main star is barely discernible in this nearly overhead image. The left-side horizontal arm was 14 centimeters long. Any seeding object for the blue-green star at top is not visible. The thin, black loops surrounding stems, and on leaves in the lower third, are shadows from troughs within a wavelike pattern caused by water surface tension.

The central part of a melting star along the northern shore. The width across the horizontal center line is 14 centimeters. At center is the seed-rich twig. Dendritic mosaics were still between arms. The dark, wavy lines indicate surface tension fringes within meltwater on the ice; closed ellipses suggest tiny puddles. The white blotches are just-fallen snowflakes.

From December 6 to 8, daytime temperatures rose to the low 50s, and the middle and northern areas of the pond lost all ice. On December 9, temperatures dropped to 20 degrees by 7 a.m. That morning, stars and dendrites returned to the northern area. However, this day proved to be the pond’s last fall ice show episode. The formations had disappeared by December 11, and despite additional warm periods, the pond remained ice- and snow-covered for the rest of the winter.