On September 13, after two earlier attempts, Chad Kellogg and I reached the summit of previously unclimbed Seerdengpu, the high point above the heads of the Changping and Shuangqiao valleys. The north and west faces above the Shuangqiao are 1,500m granite walls, and remarkably accessible. One can take a public bus along a paved road to within a couple of hours walk of the base. The more remote south and east faces above the Changping are 1,200m high and provide a combination of big walls and alpine mixed terrain. Chad and I had made a reconnaissance of this peak in 2008, after our ascent of Siguniang (AAJ 2009).

Seedengpu has been translated as “Barbarian,” “Yeti,” or “Savage Peak.” Its northern aspect distinctly resembles the head of a savage, and our original plan was to climb the unmistakable 1,400m nose. This compelling line has attracted around a dozen expeditions over the last decade, and when we entered the park we learned that three teams were attempting it during August and September. The first of these—Japanese—had been rescued after sustaining injuries from rockfall, one week prior to our arrival.

Our itinerary called for acclimatizing in the adjacent Changping Valley until mid-September, and we promptly set to work establishing an advanced base camp at the head of the valley under the east face of Seerdengpu.

As we did so, our liaison officer informed us that a Polish team had reached the base of the north face. The abandoned Japanese tent was still visible at one-third height on the nose. Storms arrived during the first week of September and deposited the first autumn snowfall. This new snow, and crowds on the north face, encouraged us to remain focused on the mixed terrain of the east side.

On our first attempt we found solid granite climbing, free at 5.10 for the first 250m. At nightfall, we rapped back to our high camp, left a line fixed over the crux slabs, and set the alarms for a pre-dawn start. The skies deteriorated throughout the night, and by 4 a.m. 30cm of snow had fallen.

Four days later we returned in marginal weather, and as we gained the northeast ridge proper, Chad took the lead and was pleasantly surprised to find straightforward passage in a hidden gully, offering 300m of snow and mixed climbing up to M5.

At 5,200m the gully terminated at a small col, above which a series of complex gendarmes guarded the upper mountain. In the waning daylight, I led up a steep gendarme, dry-tooling a thin crack. While trying to clear snow and find some gear, my tool ripped and sent me hurling backwards for my first real alpine whipper. A bit shaken up, discouraged, and without bivouac gear, we decided to rap and try the following day. Again we awoke to new snow, so we cached food and fuel and headed down the slippery talus to base camp.

After we spent a rest day, my wife Jenna called in a splitter forecast, so we set off at 1:30 p.m. for the 25km of swamp and talus leading to the base of the route. We agreed that we didn’t want to risk getting stopped by a sudden storm again, and thought the best strategy would simply be to begin climbing as soon as we reached the base. We started up the route at 11:30 p.m., and with the 5.10 slabs running with snow melt, I was forced to do interesting A2 by headlamp. We had left a line fixed over a section of the second pitch, but, nervous about the rope’s integrity after the last storm, I free-climbed most of the wet and snowy pitch “protected” by a Ropeman.

Reaching our previous high point at dawn, we traversed left beneath the gendarmes, completing four 5.10 C2 horizontal pitches to arrive at simul-climbing terrain on the upper mountain. The weather was holding, and we climbed the last 300m to the summit ridge in a single pitch. The ridge offered spectacular cornice walking and easy mixed climbing, with giant raptors flying below. At 2:30 p.m. we reached the top and enjoyed unmatched views of the entire range. For the first time I was able to mentally organize the complex topography of the Quonglai Mountains; it was my third summit in this area, but Chad’s seventh.

The descent went relatively smoothly. The traverses were difficult to reverse, but we sorted them with a few pendulums and sideways raps. Our lead line suffered two debilitating core shots, and during the final, overhanging rap our tag line became hopelessly stuck behind a flake, and we had to leave it behind. We arrived at high camp at 11:30 p.m. after 34 hours on the go. We named the 1,000m route Headwaters, after its obvious position in the hydrology of the region, a gesture to the Yeti himself, and as an acknowledgment of the alarming glacial recession taking place—a major threat to the extensive, crowded Chinese lowlands downriver.

A few days later, with John Dickey along, we nearly made the first ascent of a stunning granite spire of 5,086m, which is on the southern rim of the upper Changping, southeast of Potala Shan. In the dark, after 600m of absolutely classic free climbing (5.10) up the northwest ridge, I had to turn back 25m from the summit when faced with a steep, unprotectable arête and no bolt kit. During the descent we destroyed our only remaining lead line. On our return to Rilong we found members of the Polish team on the nose had also been hit by rockfall and were in a Chengdu hospital. They had been replaced by a group from China.

We extend our sincere thanks to the American Alpine Club, Mugs Stump, and Lyman Spitzer grant programs for their generous support. For us these exploratory trips to Asia would not be possible without the financial support of these programs. The expedition was also supported by the Four Sisters Film project; keep your eyes out for its impressive work.

Dylan Johnson, AAC