The Bay That
Used to Be
Solid Ice

What Vancouver Actually Charted in 1794

When Captain George Vancouver surveyed this coastline in 1794, the National Park Service's own account describes the glacier as having already receded about five miles back into what's now the bay, meaning the ice mass was past its 1750 peak and already retreating by the time it was first formally charted. Contemporary and later descriptions of the ice at its fullest extent describe a wall reportedly thousands of feet thick, stretching for miles across the inlet's mouth and inland toward the St. Elias Mountains, figures repeated widely in popular accounts of the bay though not restated in those exact terms on the Park Service's current pages. What's solidly documented either way is the direction of travel: by 1879, when John Muir arrived seeking the glacier he'd heard about, the National Park Service states it had retreated roughly forty more miles beyond where Vancouver had marked it eighty five years earlier.

The Retreat Feeds Itself

Glaciologists studying tidewater glaciers, the kind that end in open water rather than on dry land, have found that once one retreats off the shallow ridge of debris it built at its farthest point, it drops into the deep basin the glacier itself carved out over centuries. Water depth at the ice face is the single best predictor of how fast a tidewater glacier calves, and a deeper face calves faster, which thins the glacier, which lets it retreat further into even deeper water, a feedback loop that runs largely on its own momentum rather than tracking air temperature year to year. Glacier Bay is one of the clearest real-world examples of this cycle on record, and within roughly a century of Vancouver's survey, the retreat had split one enormous ice mass into a maze of separate glaciers spread across a bay more than thirty miles long. The bare ground the ice leaves behind is rising too, measured at up to thirty two millimeters a year in the upper bay, among the fastest rates of land uplift recorded anywhere on the planet today, the direct result of the crust rebounding as roughly thirty four hundred cubic kilometers of ice weight has lifted off it.

Not Every Glacier Retreated the Same Way

Muir Glacier, the one John Muir himself sought out in 1879 and the one that still carries his name, is the textbook retreat case: it pulled back roughly twenty miles between 1892 and 1980, thinned by more than half a mile in places over the following decades, and by the mid-1980s had retreated entirely out of the water, no longer a tidewater glacier at all. Not every glacier in the bay tells the same story. Johns Hopkins Glacier is one of a small number the National Park Service identifies as currently advancing rather than retreating, a genuine outlier worth watching for on a boat tour precisely because it runs against the bay's dominant trend. Today the park counts roughly a thousand glaciers total, sixty two of them named, with only seven still actively calving into tidewater, covering somewhere around thirty percent of the park's total area.

A Century of Return Visits to the Same Spot

Starting from photographs taken in 1886 by George Frederick Wright, researchers have returned to the same vantage points repeatedly, most recently through a dedicated National Park Service repeat-photography project running from 2009 to 2019, building a side-by-side visual record spanning more than a century. The Park Service describes the result as making Glacier Bay's retreat, in its own words, perhaps the best documented glacial retreat anywhere in the world. That record matters beyond this one bay: across Alaska's more than one hundred thousand glaciers, the Park Service states that ninety five percent are currently thinning, stagnating, or retreating, and the pace of that thinning is increasing, which makes Glacier Bay's century-long photographic baseline a reference point for understanding what's happening at scale everywhere else.

Facts and figures checked against the National Park Service's Glacier Bay Glacial History page, its 2025 park fact sheet, and its "On the Trail of the Glaciers" repeat-photography article. NPS sources themselves aren't fully consistent on the total retreat distance, citing both 60 and 65 miles in different current pages; we've used the more recent 60-mile fact-sheet figure. Muir's 1879 retreat distance is likewise cited as both 40 and 48 miles across sources; we've used NPS's own current 40-mile figure. Vancouver-era ice thickness estimates are widely repeated but not restated in those specific terms on NPS's current pages, and are presented here without an exact figure for that reason.

Stand at the mouth of the bay today, where Vancouver's ships would have had to stop against a wall of ice, and there's nothing there but open water and a view sixty miles deep. The ice that used to fill this exact spot is still out there, just much further up the bay, still calving, still retreating, still being photographed from the same rocky points people have stood on since 1886. Glacier Bay didn't just lose its ice. It handed scientists the clearest, longest-running before-and-after picture of that loss that exists anywhere on Earth.