Some names and details in this story have been changed for anonymity and confidentiality. Not all photographs correspond to actual events.
The ledge at 10,800 ft. The disappearance of Daniel Marsh on Mount Reineer and the 22 days that changed everything.
The alarm on Daniel Marsh’s phone went off at 3:47 in the morning. He had set it for 3:50, but he was already awake.
Had been for nearly an hour lying on his back in the narrow guest bed of his younger sister’s house in Auburn, Washington, staring at the ceiling while the digital readout on the bedside clock crawled through its pale blue minutes.
He was 34 years old, lean as a fence post with the kind of angular face that made him look perpetually skeptical even when he wasn’t, and hands that were rough across the knuckles and smooth at the palm from years of rope and ice.
He had been climbing since he was 19. He had summited 11 mountains across four countries.
He had never once in 15 years of this particular version of living slept well the night before a major ascent.
He reached over and silenced the alarm before it rang. By 4:10 in the morning, he was downstairs in the kitchen, moving quietly so as not to wake Clare and her husband, pressing the button on the coffee maker that Clare had loaded for him the night before.
The kitchen smelled of dryer sheets and the faint residue of the pasta dinner they had shared at 6:00 the previous evening.
On the refrigerator, held in place by a magnet shaped like a Washington state ferry, was a drawing his niece Meline had made for him, a stick figure with a triangular mountain behind it.
And in the looping letters of a 7-year-old, the words, “Uncle Dan goes up the big mountain.”
He had folded it carefully and placed it in the top pocket of his approach pack before going to bed.
His gear was in the entryway. He had packed it 3 days earlier in his apartment in Portland, Oregon, and then repacked it once on Tuesday afternoon before loading it into his car.
He had a system, and the system had kept him alive on several occasions when conditions turned faster than forecasts predicted.
The pack weighed 47 lb, heavier than he wanted it, lighter than it had been on his first serious alpine climb, the north ridge of Mount Baker in the spring of 2013 when he had carried 58 lb, and arrived at the summit feeling like he had been turned inside out.
The 47 lbs included a four-season tent, [music] a minus20 sleeping bag, double mountaineering boots rated to -40 Fahrenheit, a full rack of glacier protection, 200 ft of 9 mm rope, four aluminum pickets, six ice screws, two ice axes, his primary black diamond Viper, and a backup pzil that he had used for 8 years, and 6 days of food calculated at 3,500 calories per day.
He was planning a 3-day ascent of Liberty Ridge if conditions held. Liberty Ridge is not a beginner’s route.
It is not even, by any reasonable measure, an intermediate route, rising out of the Carbon Glacier on the remote north side of Mount Reineer.
It climbs over 11,500 vertical feet in roughly 10 to 12 m, running between the sheer nightmare of Willis Wall to the east and Liberty Wall to the west, ascending a narrow, exposed spine of snow and ice.
All the way to Liberty Cap at 14,110 ft before descending via the Emmens Winthrop Glacier on the mountains opposite flank.
The National Park Service route brief for Liberty Ridge described it plainly as the hardest and most dangerous regularly climbed route on Mount Reineer.
In 2014, an avalanche on the ridge had swept away six climbers. Two of them experienced guides.
The accident left no survivors and only partial remains. Daniel Marsh had climbed Liberty Ridge once before.
In May of 2021, he had made the summit with a twoperson team, his partner Greg Holloway, leading the Krux section just above the Black Pyramid, while Daniel bellayed from a fragile stance at 12,800 ft with the wind coming off the northwest at an indicated 24 mph.
They had made liberty cap in 9 hours from Thumb Rock Camp. Daniel had stood on the summit for 11 minutes before the cold drove them down.
He had not felt triumphant. He had felt, as he always did at altitude, a species of crystalline alertness that was something like fear and something like joy and was ultimately neither.
This June, he was going back alone. His climbing partners had heard about the solo plan with varying degrees of alarm.
Greg Holloway had called from his home in Seattle and delivered a calm, methodical argument lasting 18 minutes, citing statistics and case studies from the NPS incident reports that both of them knew by heart.
His friend Kevin Marshand, who ran a climbing gym in Portland and had done Liberty Ridge the previous spring as part of a guided group, had simply said, “That’s insane, man.
That’s genuinely insane.” Daniel had listened to both of them, thanked them, and done nothing differently.
He had filed his climbing permit with the NPS at the White River Wilderness Information Center on the afternoon of Wednesday, June 11th, listing himself as a solo climber with an expected return date of Saturday, June 14th.
He had provided his emergency contact, his sister Clare Marsh Devo of Auburn, and his satellite communicator device number.
He had paid the climbing fee. He had received a briefing from the ranger on duty, a compact, weather hardened woman named Ranger Sylvia Ortega, who had looked at his permit paperwork with an expression that was professionally neutral and said, “Have you done Liberty before?”
When he said yes once, she nodded and handed him a laminated route card without further comment.
The route card had a phone number on the back for the climbing rangers and a note that read in bold print, “Liberty Ridge has claimed more lives than any other route on this mountain.
Turn around before you need to.” [music] He had pocketed the card. Now in his sister’s kitchen at 4:18 in the morning, he poured his coffee into the stainless travel mug Clare had set out for him and stood at the window above [music] the sink, looking at the darkness of the backyard.
The motion sensor light above the garage had come on at some point. He couldn’t tell when, and it illuminated a circle of lawn, due dark and very green, with a plastic tricycle in the center that Meline had outgrown, but nobody had moved.
The light made the dew on the grass look like broken glass. He drank half the coffee.
He rinsed the mug. He left it on the drying rack. At 4:31 in the morning, he loaded his pack and his boot bag into the back of his 2019 Honda CRV and started the engine.
He had planned to leave a note for Clare, but he hadn’t been able to find paper and so instead he sent her a text message.
Heading out now. Permit filed. Return Saturday, June 14th. In reach on love you. He watched the message show as delivered.
Then he put the car in reverse and backed out of the driveway. The drive from Auburn to the White River Campground trail head took approximately 65 minutes in pre-dawn traffic.
He drove east on State Route 410, climbing steadily through the foothills as the sky lightened imperceptibly ahead of him.
The Cascade foothills dark and bulky on either side of the narrow highway. He had the radio on for the first 20 minutes.
A news station out of Seattle running a segment on a city council housing dispute.
And then he turned it off and drove in silence with the windows down 2 in smelling the cool air that was already carrying the trace of glaciated rock and high snow field.
That particular scent that existed nowhere else and that he had never been able to describe adequately to anyone who had not been there.
He arrived at the White River Campground parking lot at 5:44 in the morning. The lot was not empty.
Two other vehicles were present. A dark green Toyota Tacoma with a roof rack loaded with plastic gear boxes and a Subaru Outback with a rear window plastered in climbing and ski resort stickers.
Neither vehicle had people in or near them. He pulled into a spot, turned off the engine, and sat for a moment in the quiet.
Through the windshield, the sky above the treeine was the pale gray blue of pre-dawn in the mountains, entirely clear with a single contrail dissolving slowly near the eastern horizon.
He checked the weather app on his phone and then cross-referenced it against the UW atmospheric sciences mountain forecast he had bookmarked.
The forecast was good, better than good. A stable high-pressure system was established over the Pacific Northwest with no meaningful disturbance predicted for the next 96 hours.
Freezing level at 10,000 ft. [music] Wind at Liberty Cap Summit forecast 12 mph from the southwest.
[music] He had climbed in worse. He had summited in worse. He got out of the car at 6:02 in the morning [music] with his 47-lb pack riding high on his hips and his trekking poles clicking on the hard pack of the campground access road.
Daniel Marsh began walking toward the glacier basin trail head at the far end of the White River campground.
He was wearing a dark blue soft shell jacket over a marino base layer, glacier glasses hanging around his neck, and the double mountaineering boots on his feet.
He always started in the boots rather than approach shoes because he hated the transition and because he had the kind of feet that didn’t care about extra weight the way some climbers did.
He moved with the particular unhurried efficiency of someone who had been doing this long enough to understand that the mountain did not reward rushing.
The campground was still mostly quiet. A crow somewhere in the trees above made a single sharp sound and then fell silent.
He did not look back at the car. At 7:10 in the morning, he passed through Glacier Basin Campground and stopped for eight minutes to eat an energy bar and drink a quarter liter of water.
He noted in the small spiralbound notebook he kept in his chest pocket the time and conditions, clear sky, temperature approximately 48° [music] F at 5,600 ft, no wind.
He logged the same information into the Garmin InReach Mini 2 attached to his left shoulder strap, a satellite communicator capable of two-way text messaging and jeeps tracking, which would ping his location to a shared map every 10 minutes.
He sent Clare a brief message at Glacier Basin. All good, D. He received her response at 7:22.
Be safe. Call me when you’re down. Mattie says the stick figure is in your pocket, so you’ll have company.
He read it twice. He put the phone back in his pocket. At 8:55 in the morning, [music] he crossed St.
Elmo’s Pass at 7,500 ft, leaving the consolidated trail behind and beginning the descent onto the Winthrop Glacier.
The sun was fully up now, and the snow surface was still firm from the overnight cold, ideal conditions for travel, the kind of morning that made everything ahead seemed not only possible, but obvious.
The north face of Mount Reineer filled the entire upper left quadrant of his field of vision.
Willis wall, a vertical black and white nightmare of hanging Saraks and frozen mud cliffs.
Liberty Ridge itself, a slender, elegant line rising between it and the slightly less terrifying ramparts of Liberty Wall.
The whole thing so vast and so indifferent that it produced in him the same quiet interior adjustment he always felt in high places.
A recalibration of scale, a reminder of proportion. He crossed the Winthre Glacier, staying alert for the lateral creasses that ran across the traverse line.
He moved efficiently, probing occasionally where the snow bridging looked suspect, keeping his ice axe in hand rather than on his pack.
[music] At 10:40 in the morning, he descended off Curtis Ridge onto the Carbon Glacier at approximately 7,600 ft.
He could see from here the toe of Liberty Ridge rising out of the glacier.
A mile and a half ahead of him, a steep cleaver of snow and volcanic rock that from this distance looked almost approachable.
At 12:15 in the afternoon, he made camp on Curtis Ridge at approximately 7,800 ft in the shelter of a rock band that broke the southwest wind.
He had covered 12 mi since the trail head. He ate a freeze-dried meal, beef stew, the kind that tasted like salty paper no matter how long you soaked it, and heated water for tea.
He sent a position ping to his inreach and a brief text to Clare. Camp 1, Curtis Ridge, good conditions, summit day, Sunday, if all holds.
He did not check the weather again because it had not changed. He was organized and systematic and precise, and there was nothing in his preparation or his behavior that afternoon that suggested any reason for concern.
He was in his tent by 7:15 in the evening. Eder. He had read 40 pages of a novel, The Crossing, by Cormarmac McCarthy, which he had been working through for the third time before turning off his headlamp at 8:40.
Outside, the wind barely existed. The sky, visible through the tent mesh, was full of stars.
He fell asleep with the certainty of someone who has done everything correctly and is prepared for what comes next.
At 3:30 in the morning of Friday, June 13th, he rose in darkness and [music] began the process of making himself ready.
He was, as far as anyone knew, alive and well and [music] moving exactly according to plan.
It was the last time any evidence would place him with certainty at an established campsite.
The last [music] inreache ping from his location was transmitted from a point on the lower section of Liberty Ridge between 9,200 and 9,400 ft at 6:47 in the morning of Friday, June 13th.
After that, there was nothing. Clare Marsh Devo had not slept. She had sat in the kitchen for the first part of the night and then moved to the living room couch where she had turned the television on without sound and watched the shapes of light cross the ceiling until dawn began outside the window.
Her husband Tom, a quiet, practical man who worked in commercial insurance and had the gift of remaining steady in the presence of crisis, had stayed up with her until 2:00 [music] in the morning and then gone to bed at her insistence, leaving his phone on the nightstand with the volume all the way up.
Daniel had been expected to make contact by Friday afternoon at the latest. He had said he would text from the summit area or from Camp Sherman on the descent.
He had not. His inreach showed the last position update from Friday morning at 6:47 on the mountains north face.
And then the pings had simply stopped. At first, she had told herself the device had died, had lost power, had been stored incorrectly against his body in a way that interrupted signal.
She had researched inReach failure modes online at midnight with the specific frantic focus of someone who needs the answer to be benign.
By 5:00 in the morning, she had run out of explanations that were benign. At 6:00 in the morning, she called the Mount Rineer National Park emergency line.
The duty ranger who took the call was Ranger Dominic Lara, a 9-year veteran of Mount Reineer search and rescue who had been involved in 17 recoveries on the mountain and who had learned without ever becoming callous how to maintain the quality of calm in his voice that families needed in the first moments of these conversations.
He recorded the information carefully. Daniel Thomas Marsh, 34 years old, solo climber. Liberty Ridge permit filed June 11th.
Expected return June 14th. Last confirmed inReach contact June 13th at 6:47 a.m. From the lower Liberty Ridge area.
He read back the inReach device ID number Clare had provided. He asked a series of methodical questions, physical description, prior experience on this route, emergency equipment carried, medical conditions.
He told Clare that they would begin preliminary reconnaissance that morning and that she should remain available by phone.
He paused for one additional moment before ending the call and said, “Mrs. Marsh Deo, these things often resolve themselves on their own, but we take every report seriously and we’re going to move on this quickly.”
She said, “He has his sister’s niec’s drawing in his pocket.” She didn’t know why she said it.
Ranger Lara noted it in the call log without comment. By 8:45 in the morning of Saturday, June 14th, a six-person ground team assembled at the White River Campground under the coordination of lead climbing ranger Marcus Hol, 41 years old, a former Army Mountain Infantry Specialist who had been with the NPS climbing program for 11 years, and who moved through the logistics of a search operation with the unhurried competence of someone for whom these situations had long since ceased to be emergencies and had instead become become the specific kind of work that required everything you had.
Hol briefed his team at the tailgate of the NPS operations truck. The subject had filed for Liberty Ridge had made camp 1 on Curtis Ridge on Thursday evening and the last inreache ping placed him in the lower Liberty Ridge area on Friday morning.
The ping location was precise to within 40 ft. The device had gone silent 14 hours ago.
That ping puts him at approximately 9,200 f feet on the east side of the lower ridge.
Holt said his finger on a laminated topo map spread across the hood of the truck below Thumb Rock.
There are two scenarios where that makes sense as a last known position. One, [clears throat] he made camp in that area.
The device failed or battery ran out and he’s above Thumb Rock right now in perfectly good shape with a dead communicator.
Two, something happened in or below that location. He looked at his team. We proceed on scenario two until scenario 1 proves itself.
The ground team included Ranger Ortega, the same ranger who had processed Daniel’s permit 3 days earlier, along with four additional trained climbing rescuers.
Ranger Pete Caldwell, a compact firefighter with a red beard who had grown up in Asheford, Washington, and who knew the north side of Rineer better than any active Ranger on the team.
Ranger Yuki Tanaka, 29, who had joined the climbing program the previous season and who was widely regarded as the strongest technical climber on staff and volunteers Kyle Brderick and Anita Flores from Olympic Mountain Rescue, a volunteer SAR organization that regularly augmented NPS operations on the mountain.
The sixth team member was Jorge Sandival, a certified EMT and glacier travel specialist who carried the team’s medical kit in [music] a bright orange dry bag, lashed to the top of his pack.
At 9:20 in the morning, the ground team began ascending toward Glacier Basin. The day was clear.
Temperature at the trail head was 52° F. The forecast showed the established high pressure system holding firmly through Sunday and deteriorating Monday afternoon with the approach of a Pacific low.
This gave the search team a roughly 36-hour window of decent conditions before the mountains character would begin to change in the specific ways that made everything harder.
By 1:40 in the afternoon, the team had made Glacier Basin and was moving across the interglacier toward St.
Elmo’s Pass. Ranger Pete Caldwell had paused at the upper lip of the glacier and used a pair of S Swarovski 15×56 binoculars to sweep the visible sections of Liberty Ridge, noting no movement, no tent color, no anomaly against the white and gray terrain.
He noted in his radio log, “No visual contact from Glacier Basin approach. Ridge appears quiet from this distance.”
At 3:15 in the afternoon, the NPS helicopter, a contract Airbus H125, based out of the Puialup airport that served the three major Washington national parks during the climbing season, conducted its first aerial sweep of the Liberty Ridge corridor.
The pilot, a 46-year-old contract aviator named Stuart Gaines, who had flown mountain rescue operations in Alaska and the Cascades for 19 years, brought the aircraft in from the north and made two passes along the ridge from the Carbon Glacier up to Thumb Rock, flying as close as wind conditions allowed.
Flight technician Raina Basset, riding the jump seat with a Flur thermal imaging unit and a Canon telephoto camera, scanned the ridge systematically.
They found nothing on the first two passes. On the third pass, at 3:48 in the afternoon, Basset picked up something through the telephoto lens.
She noted it immediately in her log. Possible debris field, East aspect, Lower Ridge, approximately 9,400 ft.
Small items, possibly pack contents, spread over roughly 15 meter area. No human figure visible.
Gaines marked the GPS position and radioed the ground team. The ground team reached the marked position at 7:22 in the evening in rapidly fading light.
What they found was not a body and not a scene of obvious catastrophe. It was a collection of objects arranged by nothing more deliberate than gravity.
A single aluminum picket stamped with the initials DM in black permanent marker found half buried in the snow approximately 6 m downs slope from the ridge edge.
A small stuff sack containing a first aid kit. Its zipper broken. The contents gauze medical tape.
A folded emergency bivowwac bag scattered across the snow in a roughly linear pattern suggesting the sack had tumbled and one liner glove.
The right-hand dark gray poly pro frozen into the surface at the edge of a small impact crater in the snow that was approximately 10 cm deep [music] and 20 across.
The size and shape that would be left by a boot striking the surface in an uncontrolled descent.
There was no blood. There was no body. There were no additional tracks leading away from the area in any direction that the team could conclusively follow.
In the late daylight, Ranger Marcus Holt stood at the edge of the debris field for a long moment without speaking.
He had seen this pattern before, not frequently, but enough times [music] to know what it usually meant.
The pattern of scattered objects, the glove, the picket, the stuff sack, described a brief, violent, and uncontrolled passage.
Something or someone had come down this slope fast and then disappeared over the edge of the terrain feature that began approximately 15 m above the debris field, the eastern lip of a cliff band that dropped away into the carbon glacier roughly 400 to 500 ft below.
There was no sign that the person had stopped themselves. Ranger Yuki Tanaka crawled to the lip of the cliff on her stomach and looked over.
[music] In the gathering darkness, the cliff face dropped almost vertically into a jumble of broken ice and rock debris at the glacier’s edge.
The kind of terrain that could kill a person [music] five separate ways. She could not see the bottom clearly.
She noted in her log cliff edge at marked position approximately 10 m wide, vertical to overhanging below lip, bottom not visible in current conditions.
The team bivowwacked at 8,800 [music] ft that night in two tents dug into a protected hollow below the eastern ridge.
The temperature dropped to 17° F. The stars were so clear and close that Ranger Caldwell, lying in his sleeping bag with his headlamp off, told volunteer Kyle Brick that on a night like this, it was impossible to imagine the mountain as anything other than beautiful.
“It’s almost always beautiful right before it kills you,” Brick said. And neither of them said anything after that.
The aerial team returned at 7:30 in the morning and conducted a thermal sweep of the cliff face and the carbon glacier below the debris field.
The LIIR unit calibrated for body temperature differential showed no heat signatures corresponding to a living person in the scanned area.
The team found no body on the glacier. At 11:20, a second sweep brought the helicopter in lower, barely 60 ft above the glacier surface while Basset scanned the cliff face [music] inch by inch through the telephoto.
The cliff was deeply articulated, numerous ledges, protrusions, and recesses visible from the air, but the overhang geometry made several sections completely invisible from above.
She noted in her log, multiple ledge features visible on cliff face between 10,200 and 10,800 ft, several obscured by overhang.
Cannot confirm clear sight lines to all ledge surfaces. The ground team spent Sunday conducting a systematic descent of the carbon glacier from the base of the cliff, searching the debris field at the glacier surface for any sign of an impact.
They found nothing. The glacier’s surface showed recent icefall debris. Chunks of fractured Serak ice the size of small cars, some fresh enough that their angles of fracture were still sharp and uncorroded by weather, but nothing that could be definitively linked to a human fall.
By Sunday evening, Hol submitted a preliminary incident report to NPS headquarters in Asheford. The report stated solo climber Daniel T.
Marsh 34 Portland O R last known position 9,200 to 9,400 ft. Liberty Ridge 0647 hours June 13th.
Debris field discovered east aspect lower ridge consistent with uncontrolled fall. Cliff band below debris field drops 400 to 500 ft to carbon glacier.
No body recovered. Subject considered missing. Presumed to have sustained a fall. Search continues. The search continued through Monday and into Tuesday, June 17th, despite the deteriorating weather that arrived Sunday night as forecast.
A Pacific low that brought clouds down to 7,000 feet, wind gusting to 38 miles per hour at ridge level, and the kind of driving wet snow that reduced visibility to 15 m and made helicopter operations impossible and ground operations genuinely dangerous.
The ground team withdrew to the White River campground on Monday afternoon. The helicopter conducted one attempted pass Monday evening when conditions [music] briefly improved, found nothing, and returned to base.
By Wednesday, June 18th, the active search had effectively concluded. The NPS issued a public statement.
Search and rescue operations for missing climber Daniel Thomas Marsh have been suspended due to weather [music] and terrain conditions.
A comprehensive search of the Liberty Ridge area, including aerial thermal imaging and glacier surface reconnaissance, failed to locate the subject.
Based on evidence found at the search site, rangers believe Mr. Marsh sustained a fall in the lower Liberty Ridge area on or about June 13th.
The case remains open. Anyone with information is encouraged to contact Mount Reneer National Park.
Claire Marsh Devo reached by telephone at her home in Auburn at 3:45 in the afternoon, listened to the notification without speaking for approximately 6 seconds.
Then she said, “You’re telling me he’s gone.” The NPS victim liaison, a softspoken woman named Patricia Okcoy, said, “We’re telling you we can’t find him right now and that we’re keeping the case open.”
Claire said, “That’s not what I asked.” There was a long silence. Then she said he had his niece’s drawing in his pocket.
Patricia Okcoy wrote that down. Later, she could not have said exactly why. The clue that would prove decisive, though no one knew it yet, was a single notation buried in Raina Basset’s flight log from the Sunday aerial sweep.
Multiple ledge features visible on cliff face between 10,200 and 10,800 ft. Several obscured by overhang.
Cannot confirm clear sight lines to all ledge surfaces. That notation sat in an NPS operations file in Asheford, Washington, unread, except by the people who had written it.
For 22 days, the mountain kept its silence. What follows is not precisely years of silence because the silence in the case of Daniel Marsh lasted 22 days, a number small enough to fit in a single month and large enough to contain the complete restructuring of three lives and the beginning of the restructuring of several more.
But 3 weeks at altitude [music] in the terrain in which Daniel Marsh had last been known to be moving was not 3 weeks in any ordinary human sense of duration.
3 weeks on Liberty Ridge in the shadow of Willis Wall with the carbon glacier grinding below and the weather cycling through the mountains characteristic brutality.
Temperatures dropping to -10 on the upper ridge. Wind chill at the exposed cliff bands driving the felt temperature past -30 was a different kind of time.
It was geological time compressed into the human span. And the people waiting at the base of it felt that compression in their bodies.
The way you feel pressure underwater. Claire Marsh Devo did not leave Auburn. Her husband Tom had suggested gently and repeatedly that she might want to stay with her parents in Spokane.
Her mother had driven over the first weekend after Daniel’s disappearance and stayed 4 days sleeping in the guest bed that Daniel had slept in the night before his climb, which struck everyone as either comforting or terrible, depending on when you thought about it.
Clare had declined to go to Spokane. She had declined also to return to her job as a dental hygienist at a practice in Kent, citing personal leave.
She had sat at the kitchen table and done what people do in these circumstances.
She had researched. She had contacted Alpine Search and Rescue organizations, had corresponded at length with two different private search coordinators who specialized in wilderness missing persons, and had reached out by email and phone to every climber she could identify who had recently done Liberty Ridge or was planning to do it, asking them to look to remember if they had seen anything, to report anything anomalous to the NPS, and also to her directly.
She had put together a printed information sheet with Daniel’s [music] photograph, a good photograph taken by Greg Holloway at the Summit of Adams 3 years earlier.
Daniel squinting against the wind with a very slight smile, wearing his orange hard shell and the basic facts of his disappearance.
And she had distributed it at climbing permit stations, gear shops, and ranger stations throughout the Mount Reneer area.
The photograph showed the kind of person Daniel was in the way that good photographs sometimes do.
Competent, self-contained, at ease in a landscape that was trying to kill him. Meline, who was seven and very perceptive in the way that seven-year-olds are perceptive, which is to say accurately, without the adult capacity for strategic softening, had asked her mother twice in the first week where Uncle Dan was.
And both times Clare had said, “Still on the mountain,” in a voice that Meline had processed silently.
The way children process information they understand to be serious and sad. On the sixth day, Meline had made another drawing.
This one showed the mountain again, but with the stick figure on a ledge partway up with a small speech bubble coming from the figure’s head that said, “I am okay.”
She had placed the drawing on the refrigerator next to where the previous one had been.
Nobody moved it. Greg Holloway, who lived in Seattle and who had climbed with Daniel for 11 years, and who was still capable of being direct about difficult things in the way that long friendships sometimes allow, had driven to Auburn on the third day after the search suspension and sat with Clare in the kitchen for 4 hours.
He had not said the thing that everyone else was being careful not to say.
He had said it once quietly while Tom was out of the room. Claire, the cliff face is 400 ft.
Nobody survives a fall like that onto glacier ice. And Clare had looked at him for a moment and said they didn’t find him.
Greg had said they looked. And Clare had said they couldn’t see all the ledges.
I read the flight report. There’s a notation about it. Greg had been quiet for a moment.
And then he had said okay. He had meant it as acknowledgment, not as agreement, but he had said it.
And Clare had heard it the way she needed to hear it. Kevin Marshand, the climbing gym owner in Portland, had taken a different approach.
He had organized a group of six climbers, all experienced alpine mountaineers with Liberty Ridge or equivalent credentials, to mount a secondary reconnaissance of the cliff band during a weather window in the third week of June.
The NPS had not authorized a new formal search, having determined the conditions did not warrant the risk to rescue personnel, but there was no prohibition on private climbers visiting the area.
Marshand and his team had reached the debris field on June 24th in good conditions and spent six hours on the cliff band repelling sections of the face and scanning as many of the ledge features [music] as they could safely access.
They had reached ledges at approximately 9,600 ft and 9,800 [music] ft both empty above 10,000 ft.
The geometry of the cliff made further repelling impossible without fixed anchors they had not placed and conditions on the upper face included continuous small icefall from the Sak formations above that made extended exposure dangerous.
They had descended without finding anything. Marshaw had called Clare that evening from the White River campground, and the call had lasted 3 minutes, and Clare had said, “Thank you.”
At the end of it with the kind of flatness in her voice that came from having prepared herself for the possibility of news and receiving only the continuation [music] of silence.
Her parents came back the last week of June. Her mother, a retired school teacher named Margaret Marsh, who wore reading glasses on a beaded chain around her neck, and who had the quality that some older women have of making any room feel more anchored simply by being in it, stayed in the guest room and washed dishes and made meals and did not say things that were unhelpful.
Her father, Edward Marsh, 67, a retired civil engineer with the measured bearing of someone who had spent his career making sure structures did not fall, sat in Tom’s armchair in the evenings and said very little.
On the third evening of their stay, he said to Clare, he knew exactly what he was getting into.
He always knew. He meant it as something other than comfort, though it was not uncomfort either.
It was perhaps the closest to an honest accounting of Daniel Marsh that anyone offered in those three weeks.
A man who understood the stakes precisely and had made his calculation freely. The NPS case log for the Marsh disappearance was transferred to cold file status on June 23rd, 11 days after Daniel’s disappearance.
This was not an unusual administrative designation. It meant the case was not closed, but that active resources were no longer assigned.
If new evidence emerged, a body on the glacier, a report from another climber, a satellite image anomaly, the file would be reactivated.
Until then, it would sit on a server in Asheford behind its case number 2025 M0614-m, accumulating no new entries.
On the night of June 28th, 2 weeks and one day after Daniel Marsh had last been confirmed moving on the mountain, a thunderstorm developed off the coast and moved inland over the Cascades, dropping 2 in of new snow above 8,000 ft and clearing by the following morning to the kind of washed, brilliant, improbable blue sky that the Pacific Northwest sometimes offers as a form of apology for everything else.
Mount Reineer stood against that sky without acknowledgement. It was, as it always was after weather, entirely self-contained.
The carbon glacier below Liberty Ridge accumulated its 2 in of new snow, and the cliff band between 10,200 and 10,800 ft settled under its fresh white covering, and the mountains north face looked from any angle where you might stand to observe it exactly as it always had, enormous, indifferent, and closed to inquiry.
And somewhere on a ledge that no search team had been able to reach in a position that no aerial survey had successfully resolved into certainty, something waited in the cold, something that the mountain had not yet chosen to reveal.
Troy Basset was not a rescue worker. He was a 28-year-old electrical engineering student at the University of Washington who built drones as a hobby and who had over the course of 3 years accumulated the kind of experience with autonomous aerial systems that the FAA had decided warranted a part 107 commercial drone operator certificate even if Basset had never operated commercially.
He was also a passionate amateur mountaineer who had climbed Reneer twice via the disappointment cleaver route and who had the previous autumn begun documenting his climbing and outdoor experiences through a small YouTube channel, Cascade Lens, as he called it, that had approximately 1,200 subscribers and was, if he was being honest with himself, more of an engineering documentation project than a genuine media enterprise.
He had come to the White River campground on July 4th for the specific purpose of flying his primary drone, a DJI Mavic 3 Enterprise modified with an aftermarket thermal imaging payload over the Liberty Ridge area to document glacier recession for a geomorphology paper he was co-writing with a faculty adviser at UW.
The project had nothing to do with the marsh disappearance, which Basset was aware of.
He had [music] seen the printed information sheets at the Paradise Visitor Center and the climbing permit office, but his interest on this particular day was purely geological.
The carbon glacier had receded measurably in recent years, and the documentary value of drone footage of the glacier surface and the cliff band above it was significant to his paper.
He launched the drone at 11:22 in the morning from a cleared area in the campground parking lot.
The air was calm, the sky was cloudless. He had filed the required NPS notice for drone operation in the national park the previous week and received written authorization which was unusual.
The NPS was generally restrictive about drone flights over sensitive areas. But the geomorphology project had been approved at the research coordinator level.
At 11:31 in the morning, while conducting a standard survey pass along the cliff band between 10,200 and 10,800 ft, the same cliff band that Raina Basset had noted in her June 15th flight log as containing multiple ledge features obscured by overhang.
Troy Basset’s drone flew into a camera angle that no helicopter had achieved during the search, coming in from directly below the main overhang and [music] angling upward to capture the underside of the cliff’s upper section.
He was watching the live feed on his tablet screen [music] and thinking primarily about glacier melt patterns.
Then he saw the figure. It was on a ledge approximately halfway up the cliff face at what he would later estimate as roughly 10,800 ft elevation.
The ledge was approximately 3 m wide and perhaps 1 1/2 m deep, recessed slightly under an overhang that would have made it completely invisible from above, which explained why the helicopter thermal sweep had returned nothing and partially visible from below only from a very specific approach angle.
The thermal imaging payload showed a heat differential against the cold rock. [music] Minimal, diffuse, inconsistent with the sharp signature of a living person, but present.
The visual camera showed a human form. Troy Basset had the presence of mind in the next 30 seconds to do three things in rapid succession.
He brought the drone to a hover at approximately 30 ft from the ledge and held position.
He activated the drone’s recording function, which he later confirmed had been running since launch, but which he double-cheed out of an instinct he could not have articulated at the time, and he picked up his phone with his free hand and called 911.
“I need Mount Reneer National Park emergency,” he said when the dispatcher answered. His voice was level, but his hand holding the phone was shaking.
I’m flying a drone over the Liberty Ridge cliff band and I found a person on a ledge at around 10,800 ft.
He’s not moving. I don’t know if he’s alive, he paused. On his tablet screen, the drone’s camera showed the form on the ledge in motionless, terrible clarity.
I think it might be the missing climber, he said. Daniel Marsh. The dispatcher connected him to the NPS operations center in Asheford at 11:34 in the morning.
The duty officer was Ranger Patricia Okcoy, the same Ooy who had spoken to Clare Marsh Deo on the day the search was suspended.
She took Basset’s jeep’s coordinates, [music] confirmed his part 107 authorization, and instructed him to hold the drone in position as long as battery life allowed while she initiated the response.
At 11:38, she called the lead SR coordinator. At 11:41, she called Ranger Marcus Hol at his residence in Enamclaw where he was spending his off rotation day.
At 11:43, she called Stuart Gaines, the contract helicopter pilot who was home in Puialup.
At 11:51, she pulled up the case file 2025- M A-S-614-m, which had sat in cold file status for 9 days, and typed in the first new entry.
It had received since June 23rd. Possible visual contact with subject clifface, Liberty Ridge, approximately [music] 10,800 ft.
Drone operator report. Response initiated. Troy Basset held his drone in position for 41 minutes before the battery level forced him to return it to the launch point.
During those 41 minutes, he watched the form on the ledge on his tablet screen and tried to determine whether it moved.
He could not tell. The helicopter reached the cliff band at 108 in the afternoon.
Gaines brought the aircraft in from below and to the left of the ledge using the approach geometry that Basset had described.
Coming up from the carbon glacier and angling across the face rather than descending from above.
Flight technician Raina Basset, no relation to Troy, who had flown the original June 15th survey, was again in the jump seat.
She had the thermal unit and the camera, and she had brought additionally a handheld amplifying radio in case the subject was conscious and responsive.
At 1:11 in the afternoon, she had a clear visual of the ledge. He was there.
Daniel Marsh was lying on his right side on the ledge with his back against the cliff face and his left arm extended outward toward the edge.
He was still wearing his blue soft shell jacket. His pack was partially visible behind him, wedged against the cliff face.
He was motionless. The thermal unit showed the faintest possible heat differential, barely above ambient rock temperature in the sheltered microclimate under the overhang, but present.
Definitively present. He’s there, Basset said into her radio. 10,800 ft. He is there. I have a faint thermal.
He may be alive. For 2 seconds, nobody said anything on the radio. Then Holt’s voice came back from the ground team which was already approaching from Glacier Basin.
“Copy that. What’s his status?” “I can’t tell.” Basset said, “He hasn’t moved, but I have a heat sig.”
The short hall extraction took 3 hours to set up and execute because the cliff geometry made a [music] direct approach by helicopter impossible without risking rotor contact with the overhang.
And the alternative was a vertical short hall in which the rescue technician descended on a 150 ft line from the hovering aircraft to reach the ledge while Gaines held position in conditions that included variable rotor wash off the cliff face.
The rescue technician was Ranger Yuki Tanaka, who had been at the cliff edge 3 weeks earlier looking over in the dark.
She was lowered to the ledge at 3:22 in the afternoon. What she found is described in her official report filed the following day with a precision that does not attempt to conceal the emotional weight of what occurred.
Subject located on ledge as reported. Male adult prone position. Right lateral facing cliff face.
Condition on arrival. Conscious hypothermic. Severely dehydrated. Altered mental status. Subject recognized rescue approach and attempted to vocalize.
[music] Voice extremely weak. Unintelligible. Subject was responsive to touch and light stimulation. Vital signs recorded.
Pulse 42. Irregular. Respiratory rate 6 per minute. Core temperature estimated via field assessment at approximately 28° C, 82° F.
Subject was wrapped in thermal [music] blankets. IV access established for fluid resuscitation. Secured to rescue harness with additional anchor backup to cliff face anchor points.
Short hall extraction to helicopter completed at 4:07 p.m. Subject’s pack was recovered from ledge.
Contents intact on subject’s person inside left breast pocket of soft shell jacket. One folded piece of paper.
Child’s drawing depicting mountain and stick figure. Daniel Marsh arrived by helicopter at St. Anthony Hospital in Gig Harbor at 5:19 in the afternoon of July 4th.
His core temperature on arrival in the emergency department was 27.8° C. His right leg had a spiral fracture of the tibia sustained.
It would later be determined at the moment of his fall. He had survived 22 days on a ledge 10,800 ft above sea level in temperatures that had dropped below -10 Fahrenheit on multiple nights with the contents of whatever he had had on his person when the fall arrested him against the ledge.
He was alive. He had been alive the entire time. The attending physician in the emergency department at St.
Anthony was Dr. Leila Foresight, a wilderness medicine specialist who had been on staff for 7 years and who had treated two previous prolonged exposure mountain survival cases.
She later described the case in a medical conference presentation as one of the most extraordinary examples of human physiological endurance I have encountered in my career and noted that several factors had contributed to Marshia’s survival.
The sheltered microclimate under the overhang, which moderated the extreme temperatures experienced at the exposed ridge above, the presence in his pack of survival equipment, including emergency bivvie, sleeping bag, and food.
[snorts] And she noted with a kind of clinical care for precision what she could only describe as a remarkable degree of physiological fortitude and mental self-regulation that we do not fully understand [music] and should not underestimate.
Claire Marsh Devo arrived at St. Anthony Hospital at 6:47 in the evening, 43 minutes after receiving the call from the NPS victim liaison.
She had driven from Auburn in conditions that she would later describe accurately as not entirely safe.
Tom drove. She sat in the passenger seat with her hands folded in her lap.
She did not cry on the drive. She cried afterward in the hospital corridor outside the intensive care unit while a nurse explained that Daniel was stable, conscious, and asking for water and McCarthy’s the crossing in that order.
The nurse said he was asking for water first, then the book. Claire said she knew which one he would have asked for first.
Then she laughed. Then she cried again. Tom put his arm around her and they stood in the corridor for a while.
Meline, who was at home with her grandmother, was told the news by telephone at 7:15 in the evening.
She asked if Uncle Dan was okay. She was told he was. She said, “I told you the stick figure was there.”
Then she asked if she could have ice cream since it was the 4th of July.
Her grandmother said yes. What happened on Liberty Ridge between 6:47 in the morning of June 13th and 3:22 in the afternoon of July 4th was reconstructed from three sources.
Daniel Marsh’s own account given initially in fragments during his first week of conscious recovery [music] and then more completely in a series of interviews conducted by NPS investigators and later by a journalist named Roberta Channing from the Seattle Times who had followed the missing person case and would go on to write the definitive long- form account of the incident, the physical evidence recovered from the ledge and the forensic analysis of the InReach device which yielded data that clarified ified with painful precision exactly when and how the communicator had failed.
The fall happened at 6:51 in the morning of Friday, June 13th. The inreach’s last ping had been at 647, placing Daniel at approximately 9,200 ft on the lower east slope of Liberty Ridge.
At 651, he was 4 minutes of climbing above that point, roughly 9,400 ft, when the event occurred.
He was traversing a section of hard alpine snow on the east side of the ridge, moving in the early morning cold when surface conditions were firmst, his krampons biting cleanly.
He had been climbing for approximately 3 hours and 20 minutes since leaving his camp on the Carbon Glacier.
By his own account, conditions had been excellent. The snow was consolidated. The visibility was unlimited.
The temperature was -5 C. He was moving [music] with full concentration and had not registered any specific precursor to what happened.
A section of wind slab, a dense cohesive layer of wind deposited snow that had bonded to the underlying surface at an angle that was later assessed by an NPS snow science consultant as marginally stable under normal load conditions and critically unstable under localized dynamic loading.
Released under his left krampon. The release was not an avalanche. It was, in the vernacular of alpine climbing, a sllo, a small, sudden failure of a limited snow layer, perhaps 2 m wide and 40 cm deep.
In normal terrain, a sloth of this size would be an inconvenience. On a 45° ice slope above a 500 ft cliff, it was sufficient to interrupt his cramp-on purchase for approximately 0.3 seconds.
He went down. He had time. He described this afterward in a manner that mountaineers and trauma specialists both recognize as consistent with high stress deceleration of perceived time to attempt a self-arrest.
He drove both axes into the surface. The primary ax hit the hard ice beneath the sloth layer and skated off it.
The pick angle was wrong by a fraction of a degree, a fraction produced by the violence of the transition from standing to prone.
The backup axe found purchase for approximately two body lengths, tearing a groove in the ice that he could later see when the helicopter brought him past it during the extraction and then it too released and he was accelerating again and then he was at the edge of the cliff band and then he was not on the mountain surface at all.
The fall was 40 ft not 400 as the initial aerial assessment from the clifftop had implied.
The cliff band, as subsequent mapping confirmed, was not uniformly 400 ft of vertical drop, but rather a series of stacked ledge systems and overhangs with the upper section, the section visible from below, dropping only 40 ft to the first major ledge at approximately 10,800 ft.
It was this ledge, the one that Raina Basset’s flight log had noted as obscured by overhang, that arrested his fall.
He hit the ledge on his right side. The spiral tibial fracture was the primary injury along with contusions across the right hip and shoulder and a laceration on the right side of his forehead that bled significantly for the first hour and then clotted.
He did not lose consciousness, which he later described as possibly the least pleasant aspect of the whole experience.
He assessed himself for approximately 20 minutes after the impact, lying still, running his hands along his limbs in the systematic fashion of a climber who is drilled injury assessment until it is automatic.
He established that he could not stand without risking additional instability on the ledge. He established that his pack had traveled with him and was within arms reach.
He established that his right leg was badly injured and his left was functional. He established that the ledge was in the specific topographic sense survivable.
It was 3 m wide, roughly a meter and a half deep, slightly angled upward toward the cliff face, which meant it held snow rather than shedding it, and the overhang above reduced wind exposure to near zero in all but the most extreme conditions.
He could not send an inReach message because the device had been damaged in the fall.
The antenna housing had cracked against the cliff face on impact, severing the satellite connection.
He could not call for help because he had no signal for his phone at that elevation and in that sheltered position.
He could not climb back up the cliff because of the tibial fracture which made weightbearing on the right leg impossible.
He could survive or he could not. He chose to survive. The account of the 22 days that followed is in documentary terms one of the more extraordinary records of human adaptation under extreme constraint that exists in the literature of mountain survival.
What Daniel Marsh did in those 22 days was not heroic in the cinematic [music] sense.
There was no dramatic gesture, no final push against an overwhelming obstacle, no single moment of triumph.
What he did was methodical, systematic, and relentlessly rational. The application of everything he knew about physiology, temperature management, hydration, caloric conservation, and psychological self-regulation to the single project of remaining alive until something changed.
[music] He had 6 days of food at 3,500 calories per day. He reduced to a calculated 800 calories per day immediately establishing a strict rationing schedule written on the inner cover of his spiral notebook.
6 days at 3,500 calories becomes at 800 calories per day 26 days. He had 22 days to get through.
The math was sufficient. Water was not immediately a problem because the ledge held snow.
He melted snow using his MSR pocket rocket stove, which he had sheltered from wind damage in the bottom of his pack and which still had one and a half canisters of fuel.
He rationed the fuel with equal care, melting snow only during the warmest part of each day to conserve canister pressure.
The cold was the primary threat. His sleeping bag was rated to minus20 C. The ledge temperature moderated by the overhang dropped to [music] approximately minus15 C on the coldest nights he survived.
He also wrapped his injured leg in the emergency bivvie bag and used the closed cell foam sleeping pad which he had partly destroyed by cutting it into sections to insulate the ledge surface beneath him.
He wrote in his notebook. He wrote every day. The notebook recovered from his breast pocket contained 31 pages of entries in his cramped deliberate handwriting documenting conditions, calorie counts, fuel levels, observations about the weather, the glacier below, the sounds of the mountain.
He noted on day three that a tarmaggan had appeared on the ledge edge for approximately 4 minutes and shown no fear of him.
He noted on day seven that the weather had deteriorated and he had spent 18 hours inside his sleeping bag.
He noted on day 14 that his caloric reserves were lower than his initial calculation had projected due to the additional metabolic cost of the fracture and that he was reducing to 650 calories per day.
He noted on day 19 that the fuel was nearly gone and that he would begin eating snow rather than melted water the following day, accepting the hypothermic cost as preferable to dehydration.
On day 21, he wrote, “Weather excellent today, clear sky, warm, something may fly over.”
On day 22, July 4th, at approximately 11:30 in the morning, he heard the drone.
He had heard aircraft before. The helicopter had passed twice during the initial search, and he had attempted to signal with his emergency mirror both times, but the geometry of the overhang had prevented any reflected signal from reaching the aircraft’s approach angles.
This aircraft was different. It was quieter, lower. It came from below and to his left, approaching across the cliff face rather than descending from above.
By his own account, when the drone appeared at the edge of the ledge and its camera pivoted to face him, he had exactly one coherent thought before his capacity for cognitive processing failed under the weight of 3 weeks of semi- starvation dehydration, hypothermia, and pain.
The thought was, “It found me.” He raised his left arm. He did not have the strength to raise it more than a few inches from the ledge surface, but the camera was close enough to resolve it.
Troy Bassid, watching the live feed on his tablet in the parking lot of the White River campground, confirmed in his subsequent statement to the NPS that he saw the arm move.
“It moved,” he told investigators. “Small movement, left arm about 3 in.” That’s when I called 911.
The notebook recovered from Daniel’s breast pocket contained one final entry written in the early morning of July 4th in handwriting considerably shakier than the entries that preceded it.
Day 22. Mattiey’s drawing still in pocket. Have been keeping it there specifically for this.
Figured if they found me this way, I wanted them to find both of us.
The drawing says I am okay. Optimistic. Let’s find out. The investigation that followed the recovery of Daniel Marsh was not a criminal investigation because no crime had been committed.
But it was in a meaningful administrative and epistemic sense an investigation, a systematic attempt by the National Park Service, the King County Medical Examiner’s Office, the UW Wilderness Medicine Program, and Rineer National Park Search and Rescue to understand what had happened, what had been missed, and what could be done differently.
The formal incident review completed on August 15th, 2025 and filed as NPS critical incident review 2025-m A- CIR-614 was 73 pages [music] long.
Its conclusions were measured, factual, and in certain sections nearly brutal in their precision. The review found that the initial search had been conducted according to all applicable protocols and had not made any individual decision that was unreasonable given the information available at the time.
The aerial survey on June 15th had been conducted in appropriate conditions with appropriate equipment and the inability of that survey to identify the subject on the ledge was not attributable to negligence.
The ledge at 10,800 ft was, the review confirmed, completely invisible from above and visible from below, only within a narrow approach angle that the helicopter had not achieved due to the safety constraints on rotor clearance near the cliff face.
The ground team had correctly identified the debris field and correctly characterized the fall zone.
The decision to suspend active search operations on June 18th was within NPS policy guidelines given the evidence available.
The review also found that the critical failure was systemic rather than individual. The NPS did not at the time of the Marsh incident have a drone operation protocol integrated into Liberty Ridge cliff band searches.
The helicopter survey had been conducted at the limit of what manned aircraft could safely achieve.
An unmanned aerial vehicle, Kate, an unmanned EOBM may capable of flying lower and closer to cliff features than a manned aircraft and capable of achieving approach angles geometrically impossible for larger rotorcraft would have been able to inspect the ledge at 10,800 ft during the initial search window.
This finding was not merely retrospective. The review included a section of recommendations that resulted within 8 months of its filing in the adoption of a new NPS unmanned aerial vehicle protocol for cliffband and complex terrain searches at Mount Reineer and eventually at all NPS wilderness areas with analogous terrain.
Troy Basset, the UW electrical engineering student who had found Daniel Marsh with his DJI Mavic 3 Enterprise Bispach, was named as a contributing author in the acknowledgement section of the new protocol.
He was not paid for this contribution, and he did not ask to [music] be.
He finished his geomorphology paper. It was accepted for publication in a regional geology journal in November 2025.
Daniel Marsh spent 17 days in the intensive care unit at St. Anthony Hospital in Gig Harbor.
His tibial fracture required surgical fixation, a procedure performed on July 7th by orthopedic surgeon Dr.
From Marcus Webb, who noted in his surgical report that the fracture pattern was consistent with a high energy oblique impact and that the improvised immobilization Daniel had constructed on the ledge, foam pad sections wrapped with medical tape and held in place with a section of climbing webbing cut from his cordlet, had prevented secondary displacement that might otherwise have resulted in permanent neurovvascular damage.
“He knew what he was doing,” Webb said in a hospital press statement. The field stabilization was as good as what we’d teach a wilderness first responder to do.
It’s part of why he can expect a full recovery. His core temperature took 4 days to normalize.
The hypothermia had produced, as it does in extended cases, a series of metabolic disruptions that resolved slowly and nonlinearly.
His cardiac rhythm was irregular for 6 days. His renal function was impaired for 9 days.
His cognitive function, specifically short-term recall and executive processing, [music] was measurably reduced for approximately 3 weeks postreovery.
His body weight at the time of extraction [music] was 58 lb below his pre-climb weight.
He had gone into the mountain at 174 lb and come off it at 116.
He later described the sensation of weighing that little as like being made of something different than you were before.
Ranger Yuki Tanaka who had been the first person to reach him on the ledge did zin first to reach visited him in the hospital on July 12th 8 days after the extraction.
She was not on duty. She had driven from the White River campground in her personal vehicle during her off-roation day.
She brought him a copy of a recent NPS route conditions report for Liberty Ridge.
This was, she said later, the most mountaineer specific gift she could think of and a bar of dark chocolate.
He ate the chocolate while she was still in the room in a methodical, attentive way.
The way people eat when they have not been able to eat freely for a long time.
She told him she had stood at the top of that cliff band on the night of June 14th and looked over and not [music] been able to see him.
He said he had heard something that night. She asked what. He said he wasn’t sure, voices or wind.
She said probably both. They sat for a while in the institutional quiet of the ICU room.
Before she left, she asked him carefully, not unkindly, the way you ask a question that has no tactful form, whether he would climb Liberty Ridge again.
He thought about it for a long moment. Then he said, “Not alone,” she nodded.
“That’s a good answer,” she said. Clare Marsh Dearu spent the first week of Daniel’s hospitalization at St.
Anthony, sleeping in a chair in the family waiting room and going home to Auburn only to shower and change clothes.
On the fifth day of his ICU stay, when he was fully conscious and cognitively present enough to hold a sustained conversation, she sat next to his bed and they talked for 2 hours.
She did not, as some people in these situations do, attempt to extract from him a promise never to climb again.
She knew what he was and had always known. What she said at some point near the end of the two hours was, “Three days.
You said you’d be back in 3 days.” He said, “I know. I’m sorry.” She said, “You’re not sorry.
You’re incapable of being sorry about going up there.” After a moment, he said, “That’s probably right.”
She leaned back in the chair and looked at the ceiling for a while. Then she said, “Meline’s been telling everyone you’re fine the whole time.
She knew.” He said, “I had her drawing.” “I know,” Clareire said. They did not say anything for a few minutes.
Outside the window of the hospital room, Puget Sound was visible in the middle distance.
The water dark gray green under an overcast sky. A car ferry making its slow crossing between Breton and Seattle.
The ordinary world continuing in its ordinary motions. The mountain was not visible from this window, though on a clear day it was visible from much of the region.
A white pyramid standing at the edge of visibility, present and enormous and completely indifferent to what happened in its vicinity or to any of the people it had absorbed and returned and kept.
Greg Holloway drove down from Seattle on Daniel’s ninth day in the hospital. He brought the remaining volume of McCarthy’s The Crossing, which he had found at a used bookstore on Capitol Hill.
Daniel’s copy, which had been on the ledge with him for 22 days, was damaged beyond reading by condensation and cold.
Greg sat in the same chair Clare had occupied and said very little for the first 20 minutes.
Then he said, “I told your sister you couldn’t have survived that fall. You were statistically reasonable.”
Daniel said, “I hate being statistically reasonable,” Greg said. Daniel said, “The ledge was 40 ft down, not 400.
The initial estimate from the helicopter was wrong. The overhang broke the line of sight from above.
Greg was quiet for a moment and then he said, “So, it was a geometry problem.”
Daniel considered that everything is a geometry problem. He said, “You just don’t always have time to solve it.”
The formal board of inquiry into the search operation was conducted on August 4th. Lead Ranger Marcus Holt testified for 3 hours.
He described the search operation in precise chronological detail without self-exalpation and without self- flagagillation.
When asked by the board chair whether in retrospect the search team had missed anything, he paused for a notable period before answering.
Then he said, “We missed the ledge. We couldn’t see it from the approach angles we used.
We lacked the equipment to inspect it directly.” In retrospect, a drone capable of flying below the overhang would have located the subject on the first or second day of searching.
That equipment existed. We didn’t have a protocol for using it. He paused again. He was there the whole time.
He was alive the whole time. We just couldn’t see him. Raina Basset, the flight technician who had noted in her June 15th log that certain ledge features were obscured by overhang and that she cannot confirm clear sight lines to all ledge surfaces.
Was asked whether she had communicated this concern to the ground team commander at the time.
She said she had included it in the flight log and that the log had been transmitted to the operation center.
When asked whether she had verbally flagged it as a priority, she said, “I flagged it as an observation.
I didn’t flag it as a priority because I didn’t have a way to address it.
There was no drone protocol. I didn’t have a tool to give them for solving the problem I was describing.”
She looked at the board chair steadily. I still think about it. Troy Basset gave a statement to the board that [music] was precise and brief.
He described his flight plan, his equipment, the jeep’s coordinates of the drone’s position when he acquired the visual, the time of the call to 911.
He was asked whether he had gone to the site specifically to look for the missing climber.
He said no, he had gone for his geology paper. The board chair asked whether it was coincidental then that a drone operator with exactly the right equipment happened to be in the right place on the right day.
Basset thought about this. He said, “I guess the mountain decided to give him back.
I just happened to be the tool it used.” The board chair wrote this down.
It ended up in the incident review, not in the formal findings, but in the narrative summary section, where it sat slightly awkwardly among the protocol language and the technical recommendations, like something that belonged in a different kind of document.
The NPS revised its Liberty Ridge search and rescue protocols on January 15th, 2026. [music] The revision required that any cliffband or complex terrain search on Liberty Ridge include within 24 hours of initial incident report, a drone survey using a UAV capable of flying below cliff overhangs and achieving approach angles inaccessible to manned rotorcraft.
It designated funding for the purchase of two dedicated SAR configured UAVs to be maintained at the White River campground during the climbing season.
The revision was signed by the regional director and came into effect for the 2026 climbing season.
It was the direct result of one person falling 40 ft instead of 400 and one student’s drone flying into the right geometry at the right time.
Kevin Marshand back in Portland heard about the recovery in a group text from Greg Holloway on the afternoon of July 4th.
He read it twice. He put his phone down on the counter of the climbing gym, walked to the center of the gym floor, and stood there for a minute, surrounded by the sounds of blays being tested and shoes being chocked and climbers calling instructions to each other from the walls.
Then he walked back to his office and called his own mother in San Diego, which he had not done for several weeks, and talked to her for 40 minutes about nothing in particular.
Later, he said, “I don’t know why I called her. I guess I just needed to call someone who already knew I was okay.
Daniel Marsh was discharged from St. Anthony Hospital on July 21st, 2025. He weighed 133 lb, 15 above his extraction weight, but still 41 below his pre-climb weight.
He walked out of the hospital on crutches with Clare on one side and Tom on the other.
Meline was there, having been permitted to skip one day of the summer reading program she attended.
She hugged his waist carefully because he had told her the right leg was still tender.
She said, “I knew you were okay.” He said he knew she knew. She reached into the front pocket of his discharge bag, a NPS graphic shopping bag that Ranger Ortega had inexplicably had in her truck and had sent to the hospital and felt around until she found the folded piece of paper.
She unfolded it. The drawing was stained and partially illeible from 22 days of condensation and cold, but the mountain was still visible and the stick figure was still visible and the speech bubble was still visible and inside it the words were still readable.
I am okay. She refolded it and gave it back to him. You can keep it now, she said.
He put it in his pocket. The NPS case file 2025-m O-S614-m was formally closed on July 25th, 2025.
The final entry typed by Ranger Patricia Okcoy read, “Subject recovered alive. July 4th, 2025.
Recovery attributed to UAV reconnaissance conducted by civilian operator Troy J. Basset pursuant to NPS research authorization.
All case materials transferred to critical incident review. File 2025-m a-cir-614 for protocol analysis. Subject reported in stable condition.
Case closed. She typed it, read it once, and then very briefly thought of the morning she had told Clare Marsh Devo that the search was being suspended and the silence on the other end of the line and the way Clare had said, “You’re telling me he’s gone.”
In the specific flat tone of someone who has already understood what is being said and is simply trying to make the other person say it plainly.
She thought of what she had said in return. We’re telling you we can’t find him right now.
That sentence had been true. It had always been exactly and only true. She clicked save on [music] the case file.
She closed the window. Outside the operation center in Asheford, the mountain was visible through the south-facing windows, white and enormous, and clear in the July afternoon light.
It looked, as it always looked from here, like something that had no opinion about anything that happened in its vicinity.
In his apartment in Portland, 6 days after his discharge, Daniel Marsh sat in a chair by the window with his right leg elevated on a stacked pair of climbing bags, reading the Robera Channing draft article she had emailed him for fact-checking, a pen in his hand for marginal notes.
The chair was positioned near the window that faced east, from which he could see on clear days the blue white shape of Mount Hood on the horizon, and beyond Hood on very clear days the ghost of Adams.
Mount Reneer was not visible from Portland. It was too far north and the intervening terrain was too high.
He read the article carefully. He made 12 marginal notes. All of them corrections of small factual details.
The time of the first helicopter pass, the elevation of the debris field, the temperature rating of his sleeping bag.
He did not ask Channing to change anything about the larger narrative. He was not, by his own account, a person who felt the need to control the way a story was told.
He had found in 22 days on a 3 m ledge at 10,800 ft that control was a concept with a relatively narrow range of application and that most things that actually mattered were decided by forces considerably larger than the person they were happening to.
He wrote one note in the margin next to the paragraph describing his initial discovery by the drone.
It came from below. That’s the part that keeps surprising me. Everything I know about that mountain, I know from above.
Everything that saved me came from below. He drew a small circle around the note.
Then he went back to reading. Daniel Marsh completed outpatient physical therapy for his tibial fracture in October 2025.
His orthopedic surgeon cleared him for hiking in November 2025 and for technical climbing in March 2026.
As of the time of publication, he has not returned to Mount Reineer. Troy Basset graduated from the University of Washington with a master of science in electrical engineering in June 2026.
[music] His geomorphology paper co-authored with Dr. Patricia Euan of the UW Earth and Space Sciences Department was published in the Pacific Northwest Geological Survey in the spring of 2026.
In the papers acknowledgement section, he thanks the carbon glacier which was patient. Ranger Yuki Tanaka was promoted to lead climbing ranger for the Liberty Ridge corridor in January 2026.
Her first action in the new role was to test the newly acquired NPSS configured UAV on a cliffband inspection exercise on the north face of Reneer.
The exercise was conducted in February 2026. The drone found an empty ledge at approximately 10,800 ft.
Ranger Marcus Hol retired from the NPS in April 2026. His retirement announcement posted to the Mount Reineer National Park staff page noted that he had been involved in 34 search and rescue operations during his career, 29 of which resulted in successful recovery of the subject.
He did note which cases made him think about geometry on cold nights. Meline Marsh Dearu turned 8 in August 2025.
For her birthday, she received a book about mountains. Mount Reneer remains the highest peak in the Cascade Range at 14,410 [music] ft above sea level.
It is active. It is glaciated. It is classified by the United States Geological Survey as one of the most dangerous volcanoes in the world.
Liberty Ridge [music] has a summit success rate of approximately 53% and continues to attract roughly 100 climbers per year.
The route has claimed lives in every decade since its first ascent in 1935. It has also given some of them back.
And that’s how a simple drone and a child’s faith can rewrite the laws of geometry and save a life when all hope seemed lost.
What are your thoughts on this incredible story of Daniel Marsh? Let me know in the comments below.
Don’t forget to like, subscribe for more breathtaking stories, and I’ll see you in the next
