Alpomish, Uzbekistan Highpoint via NW to NE Face

Alpomish (aka Star of the East)

Highest mountain in Uzbekistan

Survey results recognized by the peer-reviewed scientific journal Progress in Physical Geography (accepted for publication and will be online soon)

July 3-8, 2025

NW face to NE face (New route, 400m, M3 steep snow)

Eric Gilbertson and Serge Massad

July 3 – Taxi Dushanbe to Sarytoq, hike to 3000m
July 4 – hike to basecamp 4000m
July 5 – rest day
July 6 – summit, return to basecamp
July 7 – hike out to 2800m
July 8 – hike out to Sarytoq, taxi to Dushanbe

In 2023 I made the first ascent and first survey of Alpomish with Andreas and I discovered it is in fact the tallest mountain in Uzbekistan. I used six independent Abney level measurements to determine it is 25m +/- 11m taller than peak 4643, previously thought to be the country highpoint.

A 1981 Soviet topographic map represented the only previous survey of the area, and peak 4643 had been surveyed on the map but Alpomish had not. Based on my measurements, that meant Alpomish is 4668m +/-11m. This is orthometric height using the same geoid as the Soviet map, which is the BK77 geoid. (A geoid is a definition of zero elevation and different countries use different geoids).

On that trip I brought a survey grade differential GPS unit to the summit of Alpomish but it failed to acquire satellites. This meant I could only rely on my Abney level angular measurements to get a relative height of Alpomish above peak 4643. One problem with the GPS was the summit was so sharp I couldn’t mount my tripod on it. The location I could mount the tripod had poor sky visibility. This may have affected its ability to acquire satellites. Also, it was an older unit, a promark 220 from 2012.

My measurements were sufficient to prove Alpomish is the true country highpoint of Uzbekistan, but I still regretted not getting the sub-meter vertical accuracy of the summit elevation that I intended. Over the next year I refined my technical surveying methods. I got a newer Trimble DA2 differential GPS unit and I got a small flexible-leg tripod that was capable of hugging sharp summits to mount the unit directly on any summit. I practiced using it on technical peaks in Washington and generally got 2cm vertical accuracy.

In 2025 Serge Massad and Ginge Fullen were interested in bagging the Uzbekistan highpoint and I thought that would be a great opportunity to finally get a more accurate summit measurement. I wanted to climb the easiest possible route to maximize chances of success, and I thought that would probably be the northwest face.

In 2023 I’d climbed the east face, which was a 5.8, 7-pitch rock climb. I’d climbed in late August when the snow had all melted off. From the summit I’d noticed the northwest face would likely have been easier. A long ice gully led to a saddle a few pitches below the summit. The rock pitches looked doable. That route was not good in late August because it was too icy, but I suspected earlier in the season, like in June or July, it would be an easy snow climb. That would likely be the easiest route up Alpomish.

In 2024 an Uzbek climbing team led by Vladimir Dolgiy asked me for beta about Alpomish and I gave them all my information and strongly recommended that they take the northwest face route I’d discovered. They ended up making the first ascent of that route in late July that summer. They followed the snow gully up to the saddle, then went up and right on a snow face to gain the summit ridge. The true summit is then the highest gendarme on the ridge, which is the southernmost of the gendarmes, which has my green rappel anchor next to it. (I measured this to be highest using multiple Abney levels).

In 2025 I planned to take this same route, and go in early July to hopefully ensure it was snow and not ice. I would bring my survey equipment up and get a sub-meter vertical accuracy elevation measurement. I would approach from the Tajikistan side as before. There is no red tape for this approach, while approaching from the Uzbekistan side is prohibited due to landmines.

On June 30 I left Seattle and Serge and I arrived in Dushanbe, Tajikistan around 3am July 2. We checked in at the green house hostel, which I’ve used on previous expeditions to the Afghanistan highpoint and the snow leopard peaks in Tajikistan. After running a few errands we went to bed early.

July 3

Ginge landed early in the morning, coming from a diving job in St Vincent. By 7am we were loaded up in a taxi headed north. We reached Iskanderkul lake after 3 hours, stopped there for a meal, then continued to Sarytoq village. The taxi made it a mile past the village before the road got too rough.

We started hiking from there around 1130am. Interestingly, the road had been extended since 2023 and now goes all the way to a small village several miles beyond Sarytoq. There is now a solid concrete bridge across the Mura river near the village.

We stayed on the left (East) side of the river past the village following shepherd trails to the Mura-Dikondara confluence. We forded the Mura river there, then soon after made a sketchier ford of the Zombar river. Beyond that we reached a good bridge and crossed to the north side of the Dikondara river. For reference, now that there are new bridges on the Mura it is best to cross to the west side of the Mura at the last village and stay on that side.

We hiked a few km farther then stopped at 3000m for camp in a nice flat meadow.

July 4

The next morning Ginge wasn’t feeling well so decided to stay in camp. Serge and I continued up the valley, and soon forded the tributary coming down from peak 4643. A shepherd on a horse caught up to us then and stopped to say hello. We didn’t speak a common language so the conversation was pretty short.

We continued on the west side of the river in open meadows and cow trails. At the head of the valley we followed cow trails up grassy slopes on the right that led to the base of a large glacier.

Here we switched to crampons and kicked steps up to the 4000m pass. We took a break then descended the other side. This was following the same route Andreas and I had discovered in 2023. We dropped into a bowl at 3850m then made a scrambley traverse around to the next bowl to the west. In this bowl we ascended another gentle glacier to our second 4000m pass.

Here I inspected the cairn at the pass and found a note from VL Zavjalov from June 2023. I had actually emailed with him after climbing Alpomish and he’d given me interesting history about trekking in the area.

This pass gave us our first view of Alpomish in the distance. It looked intimidating, and I remember seeing that first view in 2023 and worrying whether that was the mountain we’d come to climb. (In 2023 we had no beta at all on Alpomish, not even a picture, so hadn’t been certain what it would look like.) The peak is composed of four large spires, with the southwestern most spire the highest. This spire has a 100m long N-S summit ridge with four gendarmes on it, and the southernmost gendarme is the highest.

We descended 100m down to a large heart-shaped lake just below the Uzbekistan border and set camp at a flat spot on the outflow. There were fresh bear prints in the snow nearby so we made sure to be careful with food storage.

July 5

We planned a rest day to acclimate, but I was feeling ok so I decided to scout out the approach to the climb. This would allow me to get a good timing estimate for the approach, which would help with planning our schedule for the summit push. In 2023 I’d climbed the east face, but this year we would try the northwest face, which was new to me.

I hiked around the northeast side of Alpomish and climbed a snow gully up to the lowest point on the ridge north of the peak. From there I scrambled down loose scree until reaching a chock stone. Earlier in the season this would be filled in with snow, but now it was melted out to a 10ft drop.

I managed to carefully down climb to skiers left of the stone, then continued down scree slopes to the big bowl north of Alpomish.

The northwest gully was the largest gully on the face, and luckily it appeared to be all snow. Back in late August 2023 it was all melted down to glacier ice and would have been very challenging. But it looked mellow now.

There was a bergschrund at the base but it was bridged in the left side and passable. The route looked very doable, and I noted my time of 1.5 hrs to the base. I soon turned around and retraced my route back to camp by late morning. We then rested the rest of the day.

Chris Tomer was sending us daily weather forecasts to my inreach and the next day looked perfect. Then a three day stretch of bad weather was supposed to come in.

July 6

We left camp at 4am and made good time across the glacier northeast of Alpomish. We then climbed snow up to the notch and I built an anchor for Serge to rap over the chock stone. I then down climbed to keep the anchor gear for the main climb.

We roped up at the glacier below and I led the way across the snow bridge on the left side of the bergschrund. At the rock band above I built an anchor and belayed Serge up.

From there I led up one pitch getting rock pro in the side, and belayed Serge up. The snow was pretty low angle, but the sun was starting to hit the upper face and some rocks started raining down. This was not good news.

If we pitched out the whole couloir it would probably be 7 pitches and leave us in the danger zone way too long. So we decided to simulclimb for speed.

We next did a long simulpitch, and I made sure to always have a few pieces of pro between us. As we got higher the snow started melting down to ice and I placed a few ice screws. We got to a sheltered area and regrouped. Rocks still came down occasionally.

We then did another long simulpitch of a few rope lengths to where the gully starts hooking right. That was out of the rock fall zone, so we stopped again to regroup. Above there we simuled another 100m, and the snow was completely melted down to ice in the upper half. I wished we had come earlier in the season, like June, when the gully would be a much easier snow climb.

Finally we reached the big saddle around 10am, slung a big horn, and stopped for a break. I’d taken zoomed in pictures from basecamp and knew the southeast side of the saddle was a big cornice, so we didn’t venture too close.

Above the saddle the 2024 route by Dolgiy’s team had gone up and right on snow to gain the summit ridge. That route was now melted down to ice and would be sketchy. I led out far enough to put a screw in, but decided there must be an easier and safer line. So I retreated. I noticed one of their rap anchors part way up.

Going directly up from the saddle to the summit was all rock, but pretty steep. I instead traversed left on ledges, then I saw what looked like a reasonable line going back up and right.

I belayed Serge over a short pitch, then I continued back up and right. I noticed my old rap anchor below, and we had met up with my rap line from 2023. I recalled that terrain had appeared climbable, which was good news for our chance of success.

On the next pitch there was one tricky slightly overhanging bulge to get over that was the crux move, around M3. I made the pitch short to keep the rope tight for Serge on that pitch. I kind of wished we had brought rock shoes, since that would have made the climbing easier, but we made do in crampons.

We continued up two more M2 pitches, with good belay ledges, and I saw another of my rap anchors on a ledge. Finally by noon we topped out on the summit. I added a sling to my old rap anchor on the horn next to the summit and belayed Serge up.

This was the highest gendarme on the summit ridge, as I had verified with two Abney levels in 2023. It is the southernmost gendarme, the one with my green rap anchor next to it.

I quickly got out my Trimble DA2 differential GPS unit, screwed it on the flexible tripod, and mounted it on the sharp summit spire. This tripod was a key innovation. In 2023 I’d brought a rigid tripod and could not mount it in the sharp summit spire, which contributed to me not being able to collect data. But this time I could mount it in the top. I was still nervous about it getting knocked over by the wind and falling off the cliff, so I tied it to the anchor with some cord.

We hung out for 30 minutes admiring the view and the weather was perfect, with no wind and no clouds. I could see Peak 4643 in the distance. I bet that won’t be climbed much anymore now that it is known to be the second highest peak in Uzbekistan.

At 1245pm I logged the data and we headed down. My old anchor was still in good shape, but I backed it up anyways with another green cord. We rapped straight down toward the saddle to another old green anchor of mine. From there we made one more rap to just barely reach the snow at the edge of the saddle with my single 60m rope.

Interestingly, we saw two newish pitons on this stretch with orange cord on them. They looked like possibly from 2024 or maybe 2025, but there was no evidence of other gear any higher. It’s possible another party attempted this route last summer or this spring but bailed. Indeed, going straight up would be much harder than weaving around like we did. My 2023 report has been online for a while so I wouldn’t be too surprised if another party came in for Alpomish after that.

From the edge of the snow we walked to the middle of the saddle and I slung a big horn with red cord. Here I wished we’d brought two ropes for full 60m raps, but that would have been a lot of extra weight to carry in on the long approach. So we’d have to manage with 30m raps.

We made the first rap to get below the ice back onto snow, then one more off a horn I slung. Then Serge belayed me down to a green fixed rope connected to a piton-nut anchor. It looked like maybe the 2024 team had left this and it had gotten buried in the snow. I excavated it out a bit, then cut the end and tied a loop in it. We rapped of this, then off a two-nut anchor.

All of these rappels were taking a lot of time, and by then the snow was warming up. There were occasional small slush sluffs letting loose that were troubling, and occasional rockfall. The safest move in that case was to minimize time in the danger zone. So we decided to simul down climb.

I led the way, placing occasional rock pro or ice screws so there were always at least two pieces between us. As we got lower the snow turned to thin slush on ice, and footholds were not confidence inspiring. A few more slush sluffs came down and we tried to move fast.

Finally I reached a sheltered spot just as I ran out of gear, and we regrouped there. We then made one more long simul down climbing pitch and got down below the bergschrund by 6pm. More slush sluffs continued digging out runnels in the couloir. I think another week of warm days might make that route out.

We packed up and headed back up the scree gully. Over the day an ice step at the base of the chock stone had melted out, making climbing back up more challenging. I managed to wriggle up a move of M4, which was likely to technical crux of the day! I belayed Serge up and we continued to cross the notch, then retrace our route to camp. We got back by 9pm for a 17hr day.

July 7

We slept in, then packed up and headed out by 830am. That morning clouds were already building, and we were happy to have tagged the summit in perfect weather. We retraced our route over the pass and got some good glissading in on the other side, we then scrambled around to the next bowl and up and over the last pass.

After more glissading we hiked back down to the Dikondara river, forded the tributary from Peak 4643, then made it down to our old Camp 1.

We continued down another hour, this time staying on the west side of the Dikondara and Mura rivers to avoid the sketchy crossings. By 4pm we stopped next to the Hojikishvar river at 2800m for a nice camp in the trees.

July 8

The next morning we descended to the Mura confluence, crossed the Hojikishvar at a bridge, then lower down crossed the Mura at a bridge near the village. We then followed the road down to a hotel in Sarytoq where we met Ginge. The owner was able to give us a ride back to Dushanbe that afternoon.

Survey results

I processed the static survey data with TrimbleRTX to get an ellipsoidal height. I then needed to convert to orthometric height, which is the final reported height for a mountain. To convert to orthometric height you need to use the appropriate geoid, which is a definition of zero elevation. This can be thought of approximately as mean sea level extended across land. Different countries use different local geoid models, and there exist some global geoid models. Once the geoid is known, the geoid height at the location of interest is added to the ellipsoidal height to get the final orthometric height.

The current standard global geoid is egm2008, which is a good standard for reporting country highpoint elevations if the country doesn’t have its own geoid definition.

According to the Dr. Mirmakhmudov Erkin Associate Professor, Geology and Geography faculty, Department of geodesy and cartography, National University of Uzbekistan (Erkin 2017), the BK77 geoid is the standard used on historical topographic maps of Uzbekistan, including the 1981 map including Peak 4643 and Alpomish. This means my previous measurement of 4668m +/- 11m is orthometric height using BK77 geoid. This geoid uses the Baltic Sea Level (BSL) definition of zero elevation.

There is ongoing work in Uzbekistan to convert elevations to a more accurate local geoid based on GNSS field measurements. Because the BK77 geoid was established before the use of GPS and GNSS devices, elevations using that geoid are subject to high errors when compared to the more modern geoid. Thus, a peak elevation given using the BK77 geoid may be significantly different than its elevation using a modern geoid.

I will report my elevation results using the egm2008 geoid because that is the modern global standard. They can subsequently be converted to an updated local geoid of Uzbekistan if desired. Egm2008 geoid height for any location on earth can be found using https://www.unavco.org/software/geodetic-utilities/geoid-height-calculator/geoid-height-calculator.html

I measured the orthometric height of Alpomish as 4651.37m +/-0.09m using the egm2008 geoid. Based on my Abney level measurements this means peak 4643 has an orthometric height of 4626m +/-11m using the egm2008 geoid. In this region of Uzbekistan there is an approximately 17m vertical error in the older BK77 geoid compared to the modern egm2008 geoid.

The summit coordinates of Alpomish are (38.89304,68.17954).

Note that the name Alpomish is in fact not yet officially recognized by Uzbekistan. This is just what mountaineers call the peak. One proposed official name is Star of the East. I have been told that once an Uzbek mountaineering team climbs the peak completely from the Uzbekistan side, then an official name can be given.

References

Erkin, M., Modification of the reference frame of Uzbekistan topographic maps based on the GNSS, Coordinates, 2017, available at https://mycoordinates.org/modification-of-the-reference-frame-of-uzbekistan-topographic-maps-based-on-the-gnss/

© 2025, egilbert@alum.mit.edu. All rights reserved.