“Rock!” I heard Andrew scream, and I turned to see a massive slab of schist, displaced by Andrew himself, bouncing off the wall. From my relatively safe spot, I watched Ryoji below narrowly dodge the main chunk, which hit the ground and exploded into numerous small shards that went spinning through the air. From his vantage point, Andrew asked if everyone was ok.
Ryoji was a little shaken but unhurt. We were all ok, but were on the verge of entering another steep gorge section of Malishan canyon, surrounded by unstable terrain. With rain expected in the evening, it wasn’t a matter of if we would encounter loose rock again, but when.
Taiwan’s mountains are no secret but while hiking the peaks and serrated ridgelines is increasingly popular with locals and visitors alike, the gorges that lie between them are far less explored. Japanese river tracers, known as sawanobori or shower climbers, have already ventured deep into many of the country’s gorges, relying on local knowledge to help them navigate their way upstream. But certain streams and rivers have proved to be nearly impossible to climb using their current techniques and expertise, naturally leaving them curious about what lies above and beyond.
The approach
If you can’t climb up, maybe you can go down, went our rationale. Fired up by the adventurous tales told by the sawanobori, in 2015 I joined an international team of canyoneers who were pooling knowledge with the sawanobori on an exploration of one of Taiwan’s longest and most remote gorges. The project was to be split into two legs, the first taking place over two weeks in October 2015 and the second, over three weeks in March 2016.
Originally the team decided on a never-attempted multi-day descent of Taiwan’s longest gorge, named Qia Kan. But freakish weather brought unseasonal torrential rain and cold temperatures that we knew would hamper our efforts on the planned 12-day mission. After careful consideration, we switched to Malishan river, a slightly shorter gorge featuring a smaller catchment, which we hoped could be completed before the bad weather hit.
“We all agreed that a multi-day descent still was the expedition’s goal, and Malishan would provide a minimum of eight days between approach, descent and return,” said team expedition manager Mike Harris later. “The remoteness of it was also a determining factor”.
Multi-day canyoning is still young: there are rare reports on trips lasting longer than a couple of days, including access and return, but usually long canyons are divided into upper, middle and lower sections, each allowing access and escape. Malishan was so remote that only a self-sufficient team willing to go all the way could attempt it. Establishing camps inside a canyon with steep walls isn’t the easiest of tasks, so careful planning was taken to segment the canyon, while considering potential bivouac sites.
“The dynamics change when moving with such heavy packs through a canyon” said Andrew Humphreys, the team’s technical manager, afterwards. “Your balance changes, and the added weight requires extra friction on your descender, as well as impeccable rappelling technique.”
Packing all the equipment and food we would need was a challenge on its own. Our technical approach clothes were as compact and light as possible, but we still had to carry wetsuits, harnesses, 240m of rope, a power drill and anchoring material, let alone the video and photo gear to document our descent. Hefting 55L packs, on average we were carrying 30kg each.
The crux of the approach was an intense traverse of a narrow and unstable ridge. Thick cloud and rain added to the ambiance, as one-by-one we edged across this sketchy section with our heavy backpacks. Progress was slow as we tried to get our bearings. A Tyrolean traverse was then needed in order to overcome a bluff, which set us back another two hours.
After a long 1,000-m ascent on the first day, day two of the approach presented unexpected obstacles. Our map’s accuracy was called into question when we came across three pinnacles, amid near-shear and inaccessible terrain, none of which were marked. As the day progressed, it became clear that we would not reach the head of the canyon by dusk. We were forced to establish an improvised bivouac on a ledge next to loose rock; not long before we witnessed a massive rockfall in the valley below. We began to wonder if we had made the right choice of route.
Cold, wet and with no visual references to either side, we decided to make a push towards the valley below rather than spend another night on the ridge. Finally, we came across old trail markers, which we followed to an abandoned mountain hut, our desired camp spot. It had needed a slice of luck as well as good judgement, but we had made it to the start of the canyon.
Back to back
Day 1 in the canyon set us right into canyoning mode. With a nice flow from the start, we were in for a treat.
Making use of natural features & pitons still remained the team’s priority. A nice sequence of three falls measuring 30, 70 and 100 metres brought the day to an end, just as daylight began to fade and the rain became stronger. We found a flat ledge in between two massive landslides and set up camp. Under a tarp, Ryoji Onishi, the team’s lead climber & sawanobori expert, managed to light a fire despite the wet wood. We warmed up and rejoiced. We were finally in.
To our surprise, the upper section featured extremely hard schist, no choice of natural anchors or suitable cracks for our pitons. The rigging team - those in charge of establishing anchors for us to rappel off - burned out two out of our five drill bits on the first two drops, a very poor ratio that had us worrying about further bolting capabilities for the rest of the trip.
Next morning we got a clearer view of our situation. Although not especially narrow, steep walls flanked the canyon’s sides rendering escape nearly impossible. The scale was impressive. We came across numerous tributary streams which fed the main river, and Malishan began to step up its aquatic rating, although loose rock was still the main threat.
Moving efficiently & overcoming diverse whitewater obstacles, we made good progress and arrived at the start of the third gorge, the longest of them all. It was again late to start on such a committing section, and we decided to set up camp despite the lack of an appropriate space. With loose rock above our small shelf, we would all need to sleep with our helmets on and hope for good weather overnight.
“We could possibly finish the canyon tomorrow,” Mike has said optimistically the previous night, as we analyzed the topography for the upcoming section. But although there was only 100m of vertical to lose over a kilometre-long stretch, we had underestimate the challenge of the increased water levels in the final part of the canyon. The weather wasn’t helping either, dousing us with heavy showers through the morning, slowing progress so that we arrived at the beginning of the last 4th and final gorge only at midday.
“It will be extremely difficult to cross that,” said Andrew, as we looked out from the top of the drop at a 15m waterfall ending in a giant, recirculating pool. It was decided that we should attempt to set up the anchors for this and the subsequent drop, ascend back the first two drops and set up camp once more. Akira Tanaka, one of the strongest white water canyoneers I know, made a determined effort to swim across the current, but without success. We also couldn’t set up a floating anchor - where a backpack is used to tension a rope by tossing it over the following drop, serving as a guideline for a swimmer - the pack just eddying uselessly. Perhaps a sawanobori traverse, involving a lengthy, technical climb around the pool, would be necessary?
Crux or Cave
Overcoming the crux pool was a huge task for the team. With an improvised bivouac in a cave overseeing the drop, the team discussed the potential solutions for the obstacle. The increase in water levels due to constant rain meant the current was so strong to be overcome even by strong swimmers. A double floating anchor set-up was the solution to get the team through and out just in time before the floods arrived.
With more rain the following morning, it was clear we had to finish the canyon before the water reached unmanageable levels, having already risen some 2 inches overnight. We came up with the idea of using a throw bag as the means to direct the larger rope bag (dubbed the “double floating anchor”) over the edge of the drop. Our ingenuity was immediately rewarded, succeeding at the first attempt, thus scrapping the need for the sawanobori traverse. Akira then roped up and used the guideline to pull himself across, the force of the water trying to pin him against the wall. We watched from the top as he struggled with the current. “C’mon Akira!” we all screamed, the fate of the day, and the whole trip, in his hands. After five long minutes he finally stood up at the edge of the pool, arms raised, and we could celebrate. The crux had been overcome.
Tired but knowing that it wasn’t long until we reached the end, we progress through a short section while carefully overcoming fragile & unstable rockfalls. The last of us crossed the Malishan river as the water started to turn brown from sediment brought down by the rain. We were out, and smiles and camaraderie ensued. The longest self-sufficient canyoning exploration ever was finally real. Another 800m of vertical gain through untracked terrain lay ahead to get back to our pick-up point, but we were happy to leave that for the following day.
The gruelling return hike to the top of the mountain was the last challenge before we reached the village where we had been dropped off 8 days ago. We shared the stories and remembered the obstacles, scenery and team work necessary to overcome such a canyon. Our driver met us with a smile, warm sweet tea and some snacks. We couldn’t be more grateful.
Taiwan is now a mythical destination for canyoneers seeking long and challenging descents. The job is done.
The Taiwan Canyoning Expedition was sponsored by Edelrid, Rodcle Equipment, Vade Retro, FineTrack, and Purelements.
