I wouldn’t call myself a ‘cyclist’, even though I’ve been cycle touring in China for more than a decade. I like to think of myself as someone who just uses a bike because it’s convenient, cheap and fun.
For my first trip in China I took my Brompton folding bike to ride down the remote Nu river (怒江, Nujiang) in Yunnan, reasoning that a road running alongside a river would likely be mostly level, and that going downstream I would be going downhill, on average. The reality proved to be a little more complicated, but I had a great time for two weeks cycling from the mountains of the Yunnan-Tibetan border near Bingzhongluo (丙中洛), to the sub-tropical forests and warmer climes of Liuku (六库), a stone’s throw away from the border with Burma’s Kachin state. Encouraged by this I returned to Yunnan with the Brompton a couple of years later to ride from Kunming down the Red River (红河, Honghe) to the border with Vietnam at Hekou (河口).
During the COVID-19 pandemic I turned 60 and while I dodged the infection, I lost a couple of old friends to complications of the virus. This brought on a sudden awareness of my own mortality and that I would not be able to continue hiking up mountains or pedalling down rivers in China for much longer. I decided to take early retirement at the age of 62 and embark on a longer trip to see more of China.
When pondering where to go on my big tour, I first thought about the three major rivers that run in parallel for a few hundred kilometres in southwest China and which I had already explored to some extent: the Yangtze, Mekong and Nu. It didn’t take me long to reject each of these in turn, because I had either already explored parts of them, or in the case of the Yangtze, because much of it seemed to run through parts of central China that were heavily urbanised and industrialised or obliterated by the massive Three Dams project.
Then my thoughts shifted to the Yellow River. I knew little about it because it was in northern China, which I had seldom visited. I knew that it arose in Tibetan highlands and flowed through the loess plateau regions of Gansu and Ningxia with a northward loop into the deserts and grasslands of Inner Mongolia. I was vaguely aware that lower reaches of the river running through the north China plain were considered to be the cradle of Chinese civilisation, hence the name ‘Mother River’. I also knew that the Yellow River had important links with the Silk Road to Central Asia and had been a key to the spread of Buddhism and Islam in China.
I assumed the Yellow River was well travelled and had been the subject of many books and travel journalism pieces. And yet when I went looking I found that remarkably little had been written about it in English. There were some articles about specific parts of the Yellow River, but not much about travelling the entirety of the river from its source in the Tibetan grasslands to its outlet to the Gulf of Bohai in Shandong (山东) province.
The only Yellow River travel book I found was one by an American scholar of Chinese culture, Bill Porter, who described a journey he made along the river in 1991. His book Yellow River Odyssey focused on visits to temples and cultural sites such as Taishan, reflecting his interest in classical Chinese figures such as Mencius and Lao Tzu. Bill had travelled by bus and train at a time when China’s transport and tourism infrastructure were still basic, to say the least. His descriptions of decrepit hotels and uncomfortable long bus trips reminded me of my own experiences of travelling in Yunnan in the 1990s.
In terms of walking or cycling the river, only Chinese language sources had any accounts of people who had travelled the length of the Yellow River. China media news articles from 2020 reported on a young man called Fu Xiaofeng (扶小风) who completed a ‘pilgrimage’ by walking the entire length of the river from sea to source in one year. The articles described certain sections of his epic walk, but they did not provide details or maps of his exact route.
Similarly, I found a news story about a retired 65-year old man from Ningxia called Wang Laisheng (王来生) who had cycled along the river in 2022, from source to sea in 97 days: but again, it had no details of the route he followed on his ride.
And as I looked in more detail at the maps of the Yellow River’s course I became intrigued as to how anyone could possibly walk or cycle alongside the river for its entire course. There were some parts of the river that appeared to pass through inaccessible canyons or deserts with no roads or nearby inhabited areas. The only way to cover these sections of the river would be to bypass them by travelling on the nearest public highway or tracks running parallel to the river at some distance away.
One of these inaccessible sections of the river was in Qinghai near a place I had visited a decade earlier. The Tibetan monastery town of Ragya (拉加, Lajia) is located on alpine grassland in Golok territory near the mountain of Amnye Machen. It lies along an unusual 1000 kilometre backward loop that the Yellow River makes soon after it rises around the lakes of Gyaring Tso (扎陵湖, Zhaling Hu) and Ngoring Tso (鄂陵湖, Eling Hu).
After flowing about 500 kilometres eastward, the river takes a sudden turn north at a place known as the First Bend of the Yellow River, situated at around 3500 metres altitude in the marshy grasslands of the Qinghai-Sichuan border. After this bend, the river flows to the north-west for a further 400 kilometres through a sparsely-populated highland region. It is here that the river passes through a 100 kilometre steep-sided canyon, alongside which there are no roads and few signs of human settlements. The only way to follow this part of the Yellow River would be by boat, but the river flowing through the canyon is fast, turbulent ‘white water’.
About 30 years ago a Chinese team attempted to paddle down this section of the Yellow River on inflatable rafts, but their boats sank and seven of them drowned. More recently an adventurer who calls himself 'Semit' Shen Mite (闪米特) tackled the canyon solo on a packraft and survived - but only just.
From this canyon the river eventually emerges into the loess landscapes around Longyangxia (龙羊峡) reservoir, where it again turns eastward and flows towards Lanzhou (兰州), in Gansu province, losing a lot of altitude en route.
My map suggested there were two other hard-to-access sections of the Yellow River, further downriver along the 1500 kilometre northward loop that the river makes into Inner Mongolia. The first was in the section north of Baiyin (白银) in Gansu, where the river emerges from the loess plateau into the edges of the Tengger desert. Again, I could see no cycleable road near the river for about 100 kilometres until it reaches the town of Zhongwei (中卫) to the north.
And on the southward part of this Yellow River loop into Inner Mongolia there was another section where the river passes through another canyon, this time along the border between Shaanxi (陕西) and Shanxi (山西) provinces. This was adjacent to Yan’an (延安), an area chosen for its remoteness by Mao and his Red Army general as the end point of the Long March of the 1930s. There were few towns along this part of the river and the map showed that some stretches had no road, with the main highway veering inland into the hills for up to 30 kilometres. How would walkers and cyclists follow the Yellow River at these points?
One of the fundamental questions I had to consider when planning my Yellow River cycling trip was which way to ‘do’ the river: downstream from the source of the river near the town of Maduo (玛多) in Qinghai, or upstream from the sea outlet in Shandong?
At first glance the ‘downstream’ option seemed more attractive because it would be an overall ‘downhill’ journey from the 4000 metres altitude of Maduo to literally sea level. But when I started to consider other factors such as weather and the remote locality of the Yellow River source, I decided to go with the ‘upstream’ option.
This was partly because it’s not actually possible to cycle from the ultimate source of the Yellow River, which is a small stream located in the foothills of the Bayan Har Mountains at about 4800 metres altitude on the Tibetan plateau. While there is a dirt road that runs from Maduo town for about 70 km towards the source of the river, the final section would have to be done on foot or horseback across marshy grass hills. Even if this was feasible, recent articles by Chinese visitors to the area noted that the Maduo authorities have declared the entire area around the river source to be off limits to all tourists, to avoid environmental damage to the fragile ecosystem.
The nearest place accessible by bike to the source of the Yellow River would therefore be Maduo town, which is a small truckstop on the highway between Xining and Yushu (玉树). Located at around 4,300 metres it appears to be an inhospitable place, with long, cold winters and freezing winds that extend from September to June. While the temperatures rise somewhat in the brief summer, this coincides with the onset of monsoon rains. My planned mid-April start to cycling the river would not be a good time to be travelling to Maduo. Further to that, the first few hundred kilometres of the Yellow River beyond Maduo flow through remote Tibetan grassland wilderness with no major highways.
Using Google Earth I was able to trace the route of a road track from Maduo along the river, passing occasional small settlements at places such as Huanghe (黄河乡) and Darlag (达日, Dari). The few images available showed a bleak grassland plateau lightly populated by Golok Tibetan yak herders, dotted with a few monasteries. Starting a Yellow River cycle trip at the source would also mean being ‘thrown in at the deep end’, having to contend with extreme cold weather, high altitude and the lack of facilities in a very remote location. If anything, I would rather save that for the end of the trip when I had built up some experience. The other obvious problem would be how to get to such a remote ‘start point’ with a bicycle. The nearest city is Xining, around 500 kilometres away, and that would require at least a week of cycling.
I therefore decided to start my Yellow River trip at the sea.
A glance at the map showed me that the Yellow River enters the Bohai Sea on the coast of Shandong province, at an obscure spot somewhere between the cities of Tianjin (天津) and Qingdao (青岛). The nearest city is Dongying (东营), a place I’d never heard of, and which itself was around 70 kilometres from the coast. Nevertheless, I was sure this would be more ‘do-able’, and set out on the next stage of planning for how to get to the start point and how to structure my cycling journey.
With more than 5000 kilometres of river to follow, it seemed sensible to break down the trip into more manageable sections. I soon came up with four distinct stages, based on topography, history and culture of the inhabitants.
The first section would take me across the northern coastal plain provinces of Shandong and Henan, from Dongying through cities that formed the historical cradle of Han Chinese culture: Kaifeng (开封), Zhengzhou (郑州) and Luoyang (洛阳). I also planned to make detours away from the river to places of historical significance such as Qufu (曲阜), the home of Confucius, and sacred mountains such as Taishan (泰山), Songshan (嵩山, with its Shaolin temple) and Huashan (华山). The end of this section would be in Xi’an, which while not on the river, was a place I had always wanted to visit.
![]() |
Stage 1 (click on image to enlarge) |
The second section would take me up the long northward loop of the Yellow River into the grasslands of Inner Mongolia, along a canyon that marks the boundary between Shanxi and Shaanxi provinces. The goal would be to reach Baotou (包头) and maybe take a side trip to the provincial capital of Hohhot (呼和浩特).
![]() |
Stage 2 (click on image to enlarge) |
The third section would take me back down south from Inner Mongolia along the fertile ‘Hetao Plain’ (河套) of river territory beside the Tengger desert (腾格里沙漠) and into the loess plateau country and Hui Muslim communities of Ningxia (宁夏) and Gansu (甘肃), culminating in the ancient river city of Lanzhou.
![]() |
Stage 3 (click on image to enlarge) |
The fourth and final section would be the most challenging, taking me up to the high elevations of the Qinghai plateau and navigating remote sections of the Yellow River around the major lakes and reservoirs such as Qinghai Lake and Longyangxia until I reached the headwaters in the Tibetan highlands. This section would also mean negotiating the long loop of the river through remote wilderness areas, much of it without any roads or towns en route.
![]() |
Stage 4: Original planned route in red vs actual route in blue (click on image to enlarge) |
With each section being roughly 1000 kilometres in length, and if I would be riding 50-80 kilometres of cycling a day, I estimated it would be about three weeks per stage, including rest days. I planned to be staying in hotels or hostels along the way, based on my experience from recent cycling trips in China of being able to find and book hotels easily and cheaply using the WeChat app.
On my previous bike trips in China I’d used folding bikes: either a Brompton with 16-inch wheels, or a Dahon Jetstream with 20-inch wheels. These had performed well on the road and proved capable of carrying loads of around 15 kilos. But the trips had been brief, for no more than two weeks, and had been on mostly level roads, covering 50-80 kilometres a day in the relatively mild climate of south western China. When I looked at the 5000 kilometre route along the Yellow River route going up and down hills for weeks on end and into remote areas such as Inner Mongolia, I quickly realised I would need a bike with a bit more oomph. I decided on an e-bike.
When I told friends that I was planning to use an e-bike for touring in China, some jokingly suggested this would be ‘cheating’. I didn’t see it that way. I didn’t envisage my Yellow River trip as a physical challenge or endurance test that should be done on a pedal bike. The aim of my bike tour along the Yellow River was to enjoy it, to take advantage of the slow pace of cycling and the opportunities this provides to meet people through random encounters on the road: something that doesn’t happen to people who are using cars or motorbikes. At the age of 62 and with arthritis in my right foot, using an e-bike seemed like a good way to maintain mobility while retaining the advantages of a regular bike. An e-bike would also give me the freedom to park easily in most places and even to take it into a hotel room for safekeeping.
However, while electric scooters are now everywhere in China, pedal assist e-bikes are uncommon. Local friends told me that people preferred the cheap and simple electric scooters (‘diandong che’ 电动车,) with a throttle rather than bother with pedalling.
Fortunately I was able to find a suitable folding e-bike for sale at a Dahon shop in Shenzhen, near Hong Kong. The Dahon Unio E20 was similar to the manual Dahon pedal bike I had been riding for a few years, and was able to arrange for it to be delivered to Dongying, the city in Shandong closest to my starting point at the Yellow River estuary.
That was about the extent of my planning, and in March 2025 I set off for China. After staying with my wife’s family in Guilin for a few days, I took the train to Qingdao, in Shandong province.
概括
尽管我已在中国骑行旅行十余年,却不愿自称为"骑行爱好者"。我更愿意把自己看作一个单纯因便捷、经济和趣味而选择自行车的人。首次旅行时,我带着布朗顿折叠车沿云南偏僻的怒江骑行——当时推断沿河的公路大抵平坦,且顺流而下总体算是下坡。现实证明情况略复杂,但我仍度过了愉快的两周:从滇藏交界处群山环绕的丙中洛,一路骑到与缅甸克钦邦仅一箭之遥的六库,那里有着亚热带森林与温暖气候。受此鼓舞,几年后我又带着布朗顿车重返云南,沿红河从昆明骑至中越边境的河口。
新冠疫情肆虐期间,我迈入六十岁门槛。虽未感染病毒,却有几位老友因并发症离世。这让我骤然意识到生命有限,自己在中国登山溯溪的时日无多。于是决定62岁提前退休,开启一场更漫长的中国探索之旅。
筹划路线时,我首先想到中国西南并行数百公里的三大河流——长江、湄公河与怒江,它们我都已部分探访过。但很快逐一否决:或因某些河段已走过,或因如长江般流经高度城市化、工业化的华中地区,或被巨型三峡工程彻底改变地貌。
随后思绪转向黄河。我对这条北方河流知之甚少,仅知其发源于青藏高原,流经甘宁黄土高原地区,向北拐入内蒙古沙漠草原;模糊记得下游流经的华北平原被称为中华文明摇篮,故有"母亲河"之称;也知晓黄河与中亚丝绸之路的深刻关联,以及它对佛教、伊斯兰教在中国传播的关键作用。
我原以为黄河已被广泛探索,相关游记著作汗牛充栋。然而查阅资料时惊讶地发现,英文世界关于它的记述少得可怜——虽有局部河段的文章,但几乎无人描述从青藏草原源头到山东渤海入海口的全程旅行。
唯一找到的黄河游记出自美国汉学家比尔·波特之手,记录他1991年的沿河之旅。《黄河之旅》聚焦泰山等寺庙文化遗址,体现他对孟子、老子等古贤的兴趣。彼时中国交通旅游基础设施尚简陋,比尔乘巴士与火车完成的旅程中,破旧旅馆与煎熬长途车的描述,让我想起九十年代在云南的旅行经历。
至于徒步或骑行黄河全程的记录,仅中文资料中有零星记载。2020年媒体报道过一位叫扶小风的年轻人,用一年时间完成从入海口到源头的"朝圣"徒步,但报道既未详述路线细节,也未提供路径地图。
同样,我发现2022年宁夏65岁退休老人王来生用97天完成黄河源头到入海口的骑行新闻,但同样缺乏具体路线信息。
而当我深入研究黄河河道地图时,不禁疑惑:究竟如何实现全程沿河徒步或骑行?某些河段穿过无路可通的峡谷荒漠,周边杳无人烟。要穿越这些区域,唯一方法是借助远离河道的平行公路或土路绕行。
藏传佛教寺院小镇拉加(Lajia)坐落于果洛地区的高山草原上,靠近阿尼玛卿山。它位于黄河源头从扎陵湖(Gyaring Tso)和鄂陵湖(Ngoring Tso)附近发源后,突然形成的一段罕见的、长达1000公里的逆向回弯河段上。
黄河向东流淌约500公里后,在青海与四川交界处海拔约3500米的沼泽草原上突然转向北方,此地被称为"黄河第一弯"。此后,河流向西北方向继续流经400公里人烟稀少的高原地区。正是在这里,黄河穿过一段长达100公里的陡峭峡谷,沿岸既无公路,也几乎无人烟。要沿这段黄河行进,唯一方法是乘船——但峡谷中的河水湍急汹涌,属于"白水"激流。
约30年前,一支中国探险队曾尝试用充气筏漂流这段黄河,但他们的船只倾覆,导致7人溺亡。近年,一位自称"闪米特"(Semit Shen Mite)的冒险家独自使用便携式充气艇挑战该峡谷,勉强生还。
黄河最终从这段峡谷流出,进入龙羊峡水库周边的黄土景观区,随后再次转向东流,途经甘肃兰州,沿途海拔大幅下降。
我的地图显示,黄河在向北流入内蒙古的1500公里大回弯中,还有另外两段难以通行的河段。第一段位于甘肃白银以北,黄河从黄土高原进入腾格里沙漠边缘的区域。同样,我发现沿岸约100公里内没有可供骑行的道路,直到河流北抵中卫市。
而在黄河内蒙古回弯的南向河段,河流再次穿过另一段峡谷,此次位于陕西与山西两省交界处。这里毗邻延安——上世纪30年代,毛泽东及其红军将领因其偏远而选择此地作为长征的终点。该段黄河沿岸城镇稀少,地图显示部分区域甚至没有道路,主要公路向内陆山区偏离,最远距离河道达30公里。徒步或骑行者在这些地方该如何沿黄河行进?
规划黄河骑行路线时,我不得不考虑一个根本问题:该顺流而下(从青海玛多附近的源头出发),还是逆流而上(从山东入海口启程)?
乍看之下,"顺流而下"似乎更具吸引力,因为这是一趟从海拔4000米的玛多一路"下坡"至海平面的旅程。但当我开始考虑天气和黄河源头的偏远性等因素后,我决定选择"逆流而上"。
部分原因在于,实际上无法直接从黄河的真正源头骑行出发——它位于巴颜喀拉山麓海拔约4800米的高原上,仅是一条细小溪流。虽然有一条从玛多镇延伸约70公里的土路通向源头,但最后一段必须徒步或骑马穿越沼泽草甸。即便可行,近期中国访客的文章指出,玛多当局已宣布黄河源头周边区域全面禁止游客进入,以避免破坏脆弱的生态系统。
因此,自行车可抵达的最近黄河源头地点是玛多镇——西宁至玉树公路上的一个小型卡车停靠站,海拔约4300米。这里环境严酷,冬季漫长寒冷,寒风从9月持续至次年6月。短暂夏季气温略有回升,却又逢雨季。我计划4月中旬开始的骑行时间显然不适合前往玛多。此外,玛多之后的数百公里黄河段流经荒僻的藏区草原,没有主干公路。
通过谷歌地球,我大致追踪到一条从玛多沿河的土路,途经黄河乡、达日(Darlag)等零星小定居点。有限的图像显示,这是一片荒凉的高原草场,零星居住着果洛藏族牦牛牧民,点缀着几座寺院。从源头开始骑行还意味着"一上来就直面极端挑战":高寒、缺氧、偏远地区的物资匮乏。与其如此,不如将这段留待旅程后期,待积累一定经验后再尝试。另一个现实问题是如何带着自行车抵达如此偏远的"起点"。最近的城市西宁距此约500公里,仅骑到玛多就需至少一周。
因此,我决定从入海口开始黄河之旅。
地图显示,黄河在山东海岸汇入渤海,位置介于天津与青岛之间的某处偏僻地点。最近的城市是东营——一个我从未听说过的地方,且其本身距海岸仍有约70公里。不过,我确信这更具可行性,于是着手规划如何抵达起点及骑行路线。
面对5000多公里的河流,将旅程划分为更易管理的段落显得合理。我很快根据地形、历史和居民文化划分出四个鲜明阶段:
第一阶段 将穿越山东与河南的北部沿海平原,从东营途经开封、郑州、洛阳等汉文化发源地城市。我还计划绕道至曲阜(孔子故里)及泰山、嵩山(少林寺所在地)、华山等圣山。此段终点设为西安——虽不在黄河沿岸,却是我一直向往的古城。
第二阶段 将沿黄河向北的大回弯进入内蒙古草原,穿越标记陕晋边界的峡谷,目标抵达包头,或可绕行至呼和浩特。
第三阶段 从内蒙古南返,沿黄河"河套平原"经腾格里沙漠边缘,进入宁夏回族自治区与甘肃的黄土高原区,最终抵达古城兰州。
第四阶段(最终段) 最具挑战性——攀升至青海高原的高海拔地带,穿越青海湖、龙羊峡等湖泊水库周边的偏远河段,直至青藏高原的源头。此段还需绕行黄河在无人区的漫长回弯,大部分路段无道路或城镇。
每段约1000公里,若每日骑行50-80公里(含休息日),预计每段耗时三周。我计划沿途入住酒店或青旅——基于近年在中国骑行的经验,通过微信App可轻松预订廉价住宿。
此前在中国的骑行中,我使用折叠车:16英寸轮径的布朗顿(Brompton)或20英寸轮径的大行(Dahon Jetstream)。 它们在平坦路况下表现良好,载重约15公斤。但那些旅程不超过两周,且多在气候温和的西南地区日均骑行50-80公里。面对黄河沿线5000公里起伏地形及内蒙古等偏远区域,我很快意识到需要更强动力的自行车,最终选定电动自行车。
当朋友得知我计划在中国骑电助力车旅行时,有人戏称这是"作弊"。我不以为然。我并未将黄河之旅视为必须靠脚踏完成的体能挑战,而是旨在享受过程,利用骑行的慢节奏与沿途偶遇人群的机会——这是汽车或摩托车无法提供的体验。62岁且右脚患关节炎的我,电助力车既能保持机动性,又保留普通自行车的优势,还能轻松停放甚至带入酒店房间保管。
然而,尽管电动滑板车在中国随处可见,踏板辅助电助力车却并不普及。当地朋友告诉我,人们更偏爱廉价简易的 "电动车",直接使用油门而非费力踩踏。
幸运的是,我在香港附近的深圳大行专卖店找到一款合适的折叠电助力车。大行Unio E20与我骑了多年的手动款相似,并可安排寄送至山东东营——最接近黄河入海口的起点城市。
至此,我的筹备基本完成。2025年3月,我启程前往中国。 在桂林与妻子家人小住几日后,我乘火车抵达山东青岛。
1 comment:
Looking forward to reading about this great adventure, I was impressed with your route planning and your lack of concern with detailed planning thereafter, better to worry about the details once on the road! I have a particular interested to see if you document crossing the Grand Canal and if the many historical natural and man made rerouting and flooding occurrences are highlighted.
Post a Comment