Exclusive Interview: Extraordinary Disasters Are Here, Yet We Are Still at the ‘Shennong Tasting a Hundred Herbs’ Stage

by Posted on

In early October, much of northern China was hit by record-breaking rainfall, unprecedented for this time of year. Amid citizens’ exclamations of “How much longer will this rain go on?” and “Have we really become the south?”, and as farmers stood helplessly before their crops submerged in mud and rot, bereft even of tears, we marked the 36th International Day for Disaster Risk Reduction.

The International Day for Disaster Risk Reduction on 13 October is a UN-designated commemorative day aimed at raising global awareness of disaster risk and preparedness. This year’s theme is “Invest in Resilience, Enhance Disaster Risk Reduction”, emphasising that priority must be given to preparedness before disaster strikes, rather than to remedial action after the fact.

This is common sense – if you know a disaster is coming, surely everyone would prepare in advance. The real question is: do we truly understand disasters? Disaster relief worker Hao Nan made a remark that sounds counterintuitive but rings true: “Our understanding of disaster impacts is still at the ‘Shennong tasting a hundred herbs’ stage – blind trial and error without any theoretical framework – and that is no exaggeration.”

His words are well-founded. Two months after the extraordinary rainstorm that struck Miyun, Beijing, and Xinglong, Hebei, and during this year’s National Day holiday, his organisation, Zhuoming Disaster Relief (hereinafter “Zhuoming”), led a team on a return visit to the disaster area, setting out from the severely damaged Taishitun and tracing the flood’s path in an attempt to reconstruct the entire course of the disaster.

To dissect a disaster the way a researcher dissects the mechanism by which a pathogen causes disease – reasoning both backward and forward through its occurrence and assessing risk – requires not only multidisciplinary professional analysis, but also, on a societal level, an examination of how different stakeholders understand the disaster and its impacts from their respective perspectives. Hao Nan believes this work has been conducted far from sufficiently. Moreover, as a result of climate change, extreme precipitation and floods exceeding design standards will increasingly become the norm. What should the way forward look like? We hope that the Zhuoming team’s experiences and uncertainties will reach a wider audience of those concerned with weather-related disasters.

◉ During the National Day holiday, the Zhuoming Disaster Relief team revisited the disaster-stricken areas of Xinglong County, Hebei. A landslide destroyed most of one family’s home, and they are still living in a makeshift shelter of disaster relief tents, sheet metal and wooden planks. They were fortunate: when the flood came, they managed to scramble onto a rocky peak beside their house and survived there for days. Photo by Zhuoming volunteer Xie Yijing

I. The North Becomes the South: Townships and Villages Remain Forgotten

Northern China has been deluged with dense rainfall again these past few days. As many farmers have long lamented, the north is turning into the south. With extreme rainfall becoming ever more frequent, what new challenges have you encountered in disaster management?
Hao Nan: In recent years, areas already prone to waterlogging and flooding have seen floods exceeding design standards become more frequent and more concentrated. Going forward, not only the rain-soaked south but also the north may experience such disasters more often. Rather than the north–south divide, what we focus on when confronting floods exceeding design standards is the “inter-annual variability” and the “within-year distribution” of river discharge, to gauge the difficulty of preparedness. For instance, during the extraordinary rainstorm in late July, Taishitun in Miyun District, Beijing, recorded a peak discharge of 3,000 cubic metres per second, yet a typical high-flow reading during the flood season is merely 5–10 cubic metres per second, and in the off-season the river frequently ceases to flow entirely.

In the south, rivers in humid zones with high flood risk generally do not exhibit such extreme variability – thousands or even tens of thousands of cubic metres per second during the flood season, yet still two to three thousand in ordinary times, with the banks potentially flooding every few years. However, extraordinary floods with recurrence intervals exceeding 20 to 40 years are far harder to guard against – flood-fighting experience is difficult to pass down across generations.

The rise in extreme precipitation and floods exceeding design standards is also changing how the government operates. The “call-and-confirm” protocol has become the emergency management authorities’ primary task when responding to rainstorms and floods: once a red-level rainstorm alert is issued or a flood response activated, officials must phone down level by level to grassroots communities: “A heavy rainstorm is coming – evacuate residents in high-risk areas immediately!” The first time one encounters this, it is inevitably chaotic.

As climate change pushes the 400 mm isohyet northward, the probability of extreme precipitation triggering severe and catastrophic flash floods in mountain streams and flooding of small and medium-sized rivers will continue to rise in mountainous, stream-and-valley terrain between the 400 mm and 800 mm isohyets.

In August 2024, widespread rainstorm and flooding in Jianchang County, Huludao City and surrounding areas claimed at least 11 lives. A team leader who participated in both last year’s Huludao flood relief and this year’s Beijing–Chengde border flood response told me that in terms of type, scale and on-the-ground damage, the two disasters were remarkably similar.

Can we learn from these disaster areas? Can we recognise that such events could strike anyone, and then think about what to do?

◉ Flash floods typically surge down through gullies – which are also where terraced fields and roads are most easily built. In a typical flash-flood gully terrain such as the one shown here, if a horizontal river valley in front of the village and a vertical flash-flood gully behind it both erupt with flash floods and debris flows simultaneously, residents in the struck houses have very little chance of escape. Photo by Zhuoming study-tour member Ge Mingning
After the extraordinary rainstorm in Miyun, we immediately reached out to our farming friends on the outskirts of Beijing to assess the damage. But by 8 October, another wave of dense rainfall and cold snap hit North China, and even an organisation like ours, focused on food and the autumn harvest, had begun to feel some “fatigue” when it came to disaster coverage. In your observation, which regions’ disasters should receive more attention, and do disasters also have their own “blind spots under the lamp”?
Hao Nan: As early as 2013, building on predecessors’ work, I proposed the “Five Major Imbalances“: imbalance by disaster type, by scale of disaster, between urban and rural areas, across different post-disaster phases, and between pre- and post-disaster periods. These five imbalances in how society pays attention to disasters persist to this day with no essential change – the only shift is that people used to focus more on earthquakes and now care more about weather-related disasters. Resource allocation has always correlated more with public attention than with the actual scale of the disaster. And rural areas have consistently received less attention than cities.

Over the past decade and more, the trend of climate change driving more frequent disasters has become unmistakable. But it was not until the 2021 Henan “23·7” extraordinary rainstorm and flood (commonly referred to in the media as the “720 Extraordinary Rainstorm”) that society truly received its education. In the evening of 20 July, Zhengzhou’s urban area recorded 200 mm of rainfall in a single hour. The unprecedented downpour naturally triggered severe waterlogging almost immediately, and the news of more than a dozen citizens drowning in the metro ignited public outrage. Yet far more severe casualties, occurring at the same time or even earlier, remained virtually unknown to the public.

The officially announced death toll from Zhengzhou’s rainstorm and flooding was 380. Seventy per cent of those who died were from the mountainous and hilly county-level cities on Zhengzhou’s periphery, with deaths concentrated before 1 p.m. on 20 July. The vast majority perished in mountain flash floods, geological hazards, and flooding of small and medium-sized rivers; only about one in ten died from urban pluvial flooding. And yet virtually all of the public discourse focused exclusively on cases of urban pluvial flooding.

On the other side of the Yellow River, in Xinxiang City, the extremity of rainfall in the days surrounding “7·20” was no less severe than in Zhengzhou. The death toll was lower – this area is predominantly flat, water flows more slowly, and there was time to evacuate and rescue – but it was submerged in a vast sea of water. On satellite imagery, it looked like Poyang Lake even near its low-flow season, with villages turned into isolated islands. By our preliminary estimates, including Qixian and Xunxian counties downstream in Hebi through which the floodwaters flowed, nearly one million people were trapped by floodwaters; rescue teams evacuated at least 400,000 of them. At one point, Huixian alone – a county-level city in Xinxiang – required the full evacuation of all 200,000 of its residents.

◉ Floodwaters trapped residents in the north of Xinzhuang Village, Jishui Town, Huixian, Xinxiang, for five days. Currents along the village’s main road ran swift and turbulent; residents interviewed said all ground-floor homes were completely submerged. Photo by Zhejiang Finance and Economics TV

Hao Nan: After all these years, some critical issues still go undiscussed – the hazard mechanisms and risk profiles of urban pluvial flooding are entirely different from those of mountain flash floods, geological hazards, and river flooding. The risk of fatal casualties from urban pluvial flooding is far lower than from mountain flash floods, geological hazards, and river flooding, and the long-term impacts of the latter are also far more profound. Even in Zhengzhou, with rainfall this extreme, standing water from urban pluvial flooding drained away within a day, yet villages and towns along downstream rivers were submerged in succession for over half a month.

Moreover, urban flooding is primarily rainwater, which is relatively simple to clear. Urban infrastructure is also restored within days, and residents do not typically lose their livelihoods. But torrential floodwaters carry vast quantities of sediment and debris, and wherever they pass, residents’ property is utterly destroyed.

In the extraordinary flood disaster in Qiandongnan Prefecture, Guizhou, this past June, the mud and debris left in Rongjiang town after the flood receded was one to two metres thick. More importantly, mountain floods devastate farmland, roads, and infrastructure. In the severely affected townships of Xinglong County, Hebei, which we visited this time, much of the basic farmland located in mountain gullies, as well as riverside fields and terraced orchards on hillsides, was washed away. Basic farmland destroyed by debris flows is extremely difficult to regenerate and requires formal declassification from the protected farmland quota. Farmers who lose land to disaster receive a certain subsidy, but the loss of livelihood is permanent.

Looking ahead, the trend of increasingly frequent mountain floods will not only continue but may well accelerate.

◉ Heavy rain caused a severe landslide in Gutian Town, Shanghang County, Fujian. Photo by Zhou Xuan

II. Promoting Disaster Risk Reduction: Who Can Afford Three Months of Drinking with Villagers?

This year’s theme for the International Day for Disaster Risk Reduction centres on disaster resilience. You have previously emphasised that preventing disasters requires an entire system of early-warning monitoring, evacuation drills, and so on. For instance, when rescue and relief teams were deployed to Dongting Lake in Hunan, local authorities pre-positioned coaches at villages in the path of an approaching dam-break flood; the villagers were experienced and responded swiftly. Over the years, have funding and awareness for building disaster resilience improved in China?
Hao Nan: I do not know what OCHA (the UN Office for the Coordination of Humanitarian Affairs, under whose remit emergency disaster response falls) would make of this theme. The word “resilience” is bandied about endlessly in elite discourse, yet ordinary people are not buying it. Every field, from social sciences to engineering, talks about resilience, but the concept and its understanding are entirely different in each one; it is hard to settle on a unified definition, let alone a calculation. Such broad, vague concepts make it easy to forge so-called consensus. On “disaster resilience”, I believe the risk of a specific type of disaster in a given place must be quantifiable, with a clear method of calculation or a graded assessment. If it cannot be measured, then how can we verify whether the goal of “reducing disaster risk” has been achieved?\n\n”Disaster resilience” and disaster risk reduction are not the same concept either, but that does not mean the two cannot be compatible in practice – at the very least, some surface-level dialogue is possible. As for social donations for disaster risk reduction, they have not increased in the past decade and remain very scarce, with projects few and far between.\n\nThe return on investment in disaster risk reduction is hard to measure, and it is difficult to find a scientifically reliable benchmark; moreover, the more successful the effort, the closer disaster response comes to looking like everyday life, making its impact less visible. Without visible benefits, no one invests. Public opinion more readily perceives losses that have already occurred, while losses that might have occurred are hard to quantify and are seldom taken seriously.\n\nExtraordinary disasters are, after all, low-probability events. Carry out disaster risk reduction in 1,000 low-risk communities and, within the effective time window of the measures, a serious disaster may well never strike. To improve the effectiveness of disaster risk reduction investment, one must effectively identify high-risk communities – those with a higher probability of severe disaster losses.\n\n
◉Some community development organisations have begun paying attention to disaster issues. The Deep Social Work Service Centre in Conghua District, Guangzhou, discovered that in the mountainous north of Conghua many irrigation channels had fallen into disrepair over the years, posing a flood risk, and mobilised villagers to dredge and repair them. Photo by Feng Lian
\n\n
In community disaster risk reduction, or community resilience building, are there any examples at home that stand out and are worth learning from?
Hao Nan: As for the practices currently being explored by social organisations, schools are doing relatively well – the model is fairly mature and the outcomes predictable. At the community level, particularly rural communities, there are successful individual cases, but no replicable model for scale-up has emerged yet. It is impossible to scientifically verify that a method is effective – that is, that the same method applied in the next village would still work. Many disaster risk reduction activities that reach rural areas turn out to be completely out of step, misaligned with the risks farmers actually perceive. For example, you make a community risk map with villagers and ask: if an earthquake struck, which houses in your village would be unsafe? The villagers say, “I’ve lived eighty years and never seen an earthquake. Our biggest risk is fire – plenty of houses in nearby villages have been burned.” I have been to some earthquake-stricken areas, and the truth is they are more prone to floods, with a higher flood risk; community disaster education before the quake was all about floods, and after the quake only earthquakes are discussed and floods are never mentioned again.\n\nDisaster risk reduction tends to receive its due attention only after a disaster strikes. Yet it is hard to teach people how to respond to situations they have never encountered. And it is precisely the extraordinary disasters that cause the most severe damage – historic events that exceed the local population’s frame of reference.\n\nExperience in responding to past disasters is extremely important most of the time, but when facing an extraordinary disaster it can sometimes become an obstacle instead. This year, in the disaster areas of Meizhou and Longyan in Guangdong, we observed that residents along the riverbanks had previously been in the habit of moving their ground-floor belongings and elderly relatives who had difficulty getting about upstairs when floods came. But last June’s extraordinary flood rose above the second storey…\n\n
◉(Left) In June 2025, torrential rain in the Liujiang River basin in Guangxi caused record-breaking floods in several villages; the flood level in the early hours of the 20th stood 35 cm higher than in 1996. Photo by Yang Huiguang; (Right) A Foodthink colleague joined another of Zhuoming’s post-disaster study tours in June this year, in Xiaba Town, Wuping County, Longyan, Fujian, where floodwaters reached the second storey and the high-water marks were still faintly visible. Photo by Zhou Xuan
\n\n
The foundation of working with rural communities is simply staying long enough to build trust with local people. Villagers may never have seen what you are talking about, but they are willing to believe you and cooperate. But right now, the organisations doing community work and those doing emergency disaster relief – aren’t those two separate groups?
Hao Nan: Living alongside villagers, becoming one with them – that was the working paradigm of international development organisations. In the past, the ones doing the most solid community disaster-mitigation work in China were actually a few traditional community development organisations. They realised that disaster losses were a major factor undermining development, and only then did they incorporate disaster work into their programmes. But those projects required staff to live in the village for several years. I don’t think many organisations in China can do that today, and funders are unwilling to support it either. The biggest problem right now is that we have no money to go down into the villages. Who can afford to spend three months eating and three months drinking with villagers just to communicate an effective evacuation method? It has been years since a project received funding like that.

Only the handful of organisations with long-term rural programmes can fit disaster prevention and mitigation in “as it were”. You can count them on the fingers of two hands, and they don’t necessarily understand disasters. Some organisations that are just about to start a community disaster-mitigation project will ask: Do you have PowerPoint slides? Is there a curriculum? But this is not a ready-made package you can pick up and use.

Take the act of evacuation, for instance: the goal is to move away from the hazard. But what are the properties of the risk sources in a community – a landslide, a rockfall, heavy rainfall upstream of a river – what are their dynamic mechanisms, where will the impact zone extend? Judging all this requires very specialised professional expertise. All these factors shape the timing of evacuation, the route chosen, and the contingency plan. Different disaster types demand different responses, basic community conditions vary enormously, and the various factors influencing residents’ “knowledge, belief, and action” make matters even more complex.

Many organisations are now starting to work on community climate adaptation, and they begin by assessing the community’s meteorological disaster risk. But everyone reports that this part is difficult to do. From my observation, it is very hard to integrate risk reduction or resilience improvement as a tangible project output.

Governments at all levels now place great emphasis on the protection of life and safety. Foundational work for dealing with meteorological disasters is pressed very rigorously, with responsibilities cascaded down through every tier. The meteorological, water resources, emergency management, Ministry of Natural Resources and other departments all require grassroots communities to implement disaster-avoidance measures in earnest. For example, in line with the Ministry of Natural Resources’ requirements, communities must implement “hazard point + risk zone” dual control for geological hazards, produce awareness instruction cards for geological-hazard prevention, and so on. In theory, rural communities are all supposed to set up publicly funded posts such as disaster information officers and geological-hazard risk monitors.

◉ A flash-flood defence awareness instruction card. Photo by Zhou Xuan

Hao Nan: But carrying this out is equally difficult. In rainstorm and flood disasters, the flash floods and debris flows at the hazard’s origin often arrive simultaneously with the warning itself, which is very different from what most contingency plans stipulate. Whether one can, in peacetime, devise measures that will genuinely solve problems when disaster strikes has become the dividing line for professional competence in community disaster-prevention projects. Sometimes, a community project can raise awareness, but the actual effectiveness in disaster prevention and mitigation is anyone’s guess.

An organisation once invited us to conduct a disaster risk assessment for a village. That village had a complex risk background, requiring three or four experts from different specialist fields to carry out on-site surveying and assessment. For example, identifying the front and rear edges of a potential landslide – its cracks and uplift being precursors of failure – estimating the geotechnical structure, understanding how thick and how large a mass of soil might slide, in which direction it would move, which households it would affect, and so forth. The budget for an assessment like that does not exceed 5,000 yuan, even at the most generous.

Although the sum is small, we are willing to take part in projects like that because they are genuinely useful – yet we have never had another opportunity.

There is another category of community disaster-capacity-building projects: they form part of the post-disaster recovery package and are almost always one-off. A friend recently went to Dingri County, Tibet, to run a community earthquake-resilience project. A magnitude 6.8 earthquake had struck the area earlier in the year, and project funding was available. These projects follow a uniform pattern: they are carried out only where a disaster has occurred, and only within a year of it. They are essentially a way of providing survivors with a sense of security and psychological reassurance. Two years later, no one in the locality wants to continue.

Projects that should belong to the pre-disaster phase can only be implemented after the event. Having to wait for a disaster to actually happen in order to identify high-risk communities is simply too costly.

Your study tour this time covered Xinglong County in Hebei and Miyun District in Beijing. What was grassroots work like in those affected areas in practice?
Hao Nan: In Shangshidong Township, Xinglong County, Hebei, the biggest problem facing this high-risk township is a severe shortage of personnel. The township covers an area of over 100 square kilometres, yet the entire township’s permanent population is barely over 1,100. One administrative village that was hit very hard by the disaster has a permanent population of just around 60. The youngest person in the village was elected village head, and he is already 55. There is not a single school, kindergarten, or factory in the entire township.

 

◉ The road near Shanshenmiao Village, Shangshidong Township, was temporarily destroyed by floodwater, creating a sheer drop-off. Photo compiled from online sources by the Zhuoming team
Hao Nan: With the village so ageing, once we got there we could barely make out the local accent. Young people are not around either – the only ones we met had come home for the National Day holiday to help clean up or deliver supplies; back in Beijing they do the lowest-paid jobs, in estate agency, parcel delivery or ride-hailing. In places like this, phone push notifications for warnings simply don’t work. You first have to understand how older people actually live here; otherwise there is no way to find locally adapted, effective channels of communication.

III. Stones Rolling Down the Mountain: Louder than the Village Cadres’ Calls

Zhuoming has run more than ten study tours. You have said that going on a disaster study tour is really about dissecting the types and occurrence mechanisms of disasters. In this particular extraordinary rainstorm and flood, which risks could have been anticipated in advance using existing knowledge?
Hao Nan: Flood disasters alone comprise over a hundred types, and they often coincide with a whole range of other hazards. In a specific disaster scenario, accurately assigning a taxonomic classification or recommending the right response in advance is extremely difficult. I come from a background in oral medicine, and in terms of disaster-related practice we truly are still at the ‘Shennong tasting a hundred herbs’ stage – or rather, we have not even properly dissected the typology. Calling it “medieval” would be generous. Returning to the extraordinary rainstorm in the Qingshui River basin in late July – how did the flood that caused such heavy casualties form? The rainstorm hit and water came surging down the slopes. From flash floods in the most upstream small watersheds, through flooding of small and medium rivers, to regional, basin-wide flooding – it unfolds in stages, step by step. The flow velocity, load, and dynamics of the flood are primarily related to the “gradient” of the river channel (the ratio of elevation drop to horizontal distance over a given reach, analogous to slope). As water converges, it picks up mud, sand, and stones, all of which require detailed analysis.

The complex dynamics may not be fully explicable even to specialists. But in mountainous areas, when the rain is particularly intense, residents must be able to determine where the water is coming from – knowing whether the threat is a flash flood from the hill behind the house or floodwater in the river at the front door. That determines which direction holds the chance of survival.

Understanding flood dynamics is also crucial for rescue. This time, the Taishitun elderly care centre in Miyun was primarily inundated; water rose quickly but still took one to two hours, so residents had time to move to higher ground. Whereas at Liulimiao Town in Huairou, the village official swept away by the flash flood – that process may have lasted only seconds. There was simply no chance of rescue at the time.

So disaster prevention upstream of a flood source is far more difficult. It must rely on pre-disaster assessment to determine the rainfall threshold at which a flash flood becomes destructive in a small watershed – to know what rainfall intensity will cause converging flash floods to destroy where people live – and to set the trigger at which residents must evacuate. After all, by the time the rain has truly built up, the flash flood is imminent and evacuation time has already run out. In the mountains, floodwater outruns the red alert.

In many villages, survivors we met all described hearing what sounded like “thunder rolling through the mountains” – but it was not thunder; it was the roar of flash floods and debris flows thundering down the gullies. Beside houses demolished by landslides and debris flows, you can also hear a tremendous rushing sound and feel surging air blasts. These sounds are so loud that, when disaster strikes, they drown out the village cadres’ calls to evacuate.

This is why so many grassroots cadres have been killed or injured over the past two years – while everyone else flees, they head toward the danger. I even went to the site specifically to reconstruct the process by which a village Party secretary and his companion were swept away this time – how they stepped out of the village committee office, were knocked down by the surging mud-water cascading down, and were dragged step by step into the river channel…

◉ This year’s Miyun flood mercilessly swallowed Sohu Farm, situated along the Qingshui River. Foodthink’s reporting documented the experience of farm employees swept away by the floodwaters who narrowly escaped with their lives. Photo by Sohu Farm
How, then, does one assess in advance how much rainfall a mountainous small watershed can withstand – or, in other words, its “rainfall threshold”?
Hao Nan: To determine the rainfall threshold and thus assess disaster risk, one must precisely model how much rainfall in a given small watershed will trigger a flood capable of scouring the gully and raising downstream water levels. This requires integrating data on topography, vegetation, soil, lithology, catchment area, and slope; mapping the extent of the catchment area; and, factoring in precipitation and underlying surface conditions, calculating the surface runoff conversion rate – which falls squarely within the water resources departments’ professional remit. But by institutional division of functions, the Ministry of Water Resources is more concerned with how floodwater is generated, converged, and flows; how to carry out flood storage and regulation; and how to manage risk within the scope of existing downstream water infrastructure. How communities should respond – especially to mountain flash floods at the uppermost reaches – falls outside their line of sight and beyond their remit.

In practice, it is the Ministry of Natural Resources that addresses geological hazards. In their classification, flash floods are typically subsumed under the integrated management of small watersheds in mountain areas and merged into geological hazards.

When we got to the local level, we found that in one township the Water Affairs Bureau was still conducting household-by-household disaster risk assessments. In my view, their routine remit covers only areas along the riverbanks; whether they can cover flash floods and geological hazards is questionable.

Those with flood-modelling capability do not manage small watersheds – they focus on the main river channel; those responsible for small-watershed hazards are not proficient in hydrology and hydraulics modelling. There are clearly institutional silos here that need to be broken down.

You mentioned that floodwater outruns the red alert. Setting aside scientific modelling and hazard mechanisms, there is another question ordinary people care about most: when should evacuation happen? What knowledge tools can help us make that call, and what local experiences can be distilled?
Hao Nan: First, in the extraordinary rainstorm in the Qingshui River basin, the crucial point is that the people responsible were not asleep. In the small hours, two or three in the morning, a sudden emergency notification to evacuate residents was no easy task for rural cadres in their fifties and sixties. They still managed to avert greater loss of life – the lesson to draw is that.
*Editor’s note: After 2 a.m. on 29 July, all nine villages under Shangshidong Township successively lost contact with the township authorities, as the rainstorm triggered flash floods and debris flows; all nine administrative villages across the township were cut off from power, communications, and roads, becoming islands of isolation. Cao Zuofeng, Party Secretary of Shangshidong Township, once walked 37.9 kilometres on foot, going village by village to confirm residents’ safety.

◉ Mipu Village, Shangshidong Township, lies within a narrow, elongated valley gully and was severely hit by debris flows and flash floods. On 4 October, every road leading into the village had been washed out or collapsed; vehicles had to drive along the channel carved by the flood, with the original roadbed overhead. Former cropland in the riverbed was entirely buried beneath boulders. Photo by Xie Yijing

Hao Nan: But returning to the scientific question of how to determine evacuation timing, the problem emerges. The worst-affected areas in this rainstorm and flood used the water-level thresholds for floods with return periods of five, ten, and fifty years to decide whether to evacuate, in line with the water resources department’s flood-control response plans.

But a flood’s “return period” is an engineering design standard based on probability statistics. Using this as the criterion for human evacuation is often only applicable to the banks of the downstream river channel; upstream, it may have little reference value. This time, evacuations from river-channel flooding in Xinglong County were generally successful – convergence takes time, so residents typically had a several-hour evacuation window. The fatal events were mainly mountain flash floods, debris flows, and the like; they occur earlier, closer in time to the extreme precipitation itself, so evacuating as if it were channel flooding means running out of time.

The meteorological service’s short-term rainstorm warning has the same problem – red-level rainstorm alerts are generally issued at county level, indicating 100 mm or more of rainfall over three hours, or 50 mm or more over one hour, but they are hard to operationalise for evacuation decisions at the specific township or street level. Again, they are more relevant to downstream river flooding.

This time, Xinglong County issued a red-level rainstorm alert after 2 a.m., based on rainfall across the entire county – but between 1 a.m. and 2 a.m., rainfall at Yingnanyu Village in Shangshidong Township had already reached 91 mm in just one hour. At Zhazi Gully, further upstream of Yingnanyu, flash floods had begun rampaging around 1:30 a.m. The critical threshold I described earlier – more relevant for flash-flood warning than the county-level alert – the small-watershed rainfall threshold – had been exceeded before the red alert was issued.

So whether and when to evacuate, grassroots decision-makers largely still decide based on the actual situation on the ground. The professional expertise and resources available at the front line of disaster prevention and mitigation remain rather scarce.

So do you think evacuation timing should still be calculated locally, based on the rainfall threshold of each watershed?
Hao Nan: Yes. As soon as rainfall in the upper reaches of a small watershed approaches the disaster-triggering threshold, downstream residents must be alerted and told to evacuate immediately. There is no need to wait for rainfall statistics over a wider area or for alert notifications. Some grassroots communities have already begun calculating small-watershed rainfall thresholds. In mid-July this year, we visited a village where a geological hazard two years earlier had caused significant casualties. Based on readings from flash-flood rain gauge stations, once rainfall reaches the threshold, the village loudspeaker system sounds a high-pitched alarm, telling villagers to evacuate. It had rained the night before we arrived, and the villagers evacuated overnight. When we arrived, the village was almost empty. This has become their normal routine during the flood season. In other, more complex situations, even from a professional standpoint it is almost impossible to accurately predict in advance when and where a disaster will strike and at what scale – but at the very least, local communities should be aware of the possibility of compound risks.

Take Yangjiatai Village in Liudaohe Town, Xinglong County, Hebei, which was severely affected: flash floods and debris flows surged in from two directions almost simultaneously, completely sealing off all escape routes… This was also where geological hazards caused the highest number of casualties.

Xinglong County issued its red-level rainstorm alert at just past 2 a.m. The small watershed in which Yangjiatai Village sits did not actually experience the downpour until around 3 a.m. By four or five o’clock, the extremity of the rainfall had disrupted the geotechnical equilibrium at the high ridgeline. Multiple branch gullies upstream simultaneously erupted with flash floods and debris avalanches. A large share of the flow converged in front of Groups Four and Five of Yangjiatai Village, first sweeping away several houses and courtyard walls at the bottom of the gully; a smaller share converged into the flash-flood gully behind Group Four, from where it then razed most of the houses and fields in that group.

◉ Shangshidong Township under reconstruction. Photo by Hao Nan

IV. Treating “Cavities”: You Need Research and Community Doctors Alike

Hao Nan: Overall, in responding to meteorological disasters, China still leads the world, and there is ever less experience abroad from which to learn. Most of China’s territory falls under a monsoon climate, making it one of the countries with the most weather-related disasters in the world, and the impacts of the climate crisis were felt here early and deeply. The terrain is diverse, and virtually every disaster type is represented. Mountainous areas and river valleys are densely populated. Only a handful of countries share all of these characteristics simultaneously. At present, the geological structures of every region in China, each sub-category of disaster, and the dynamic mechanisms of every stage of flood disasters are all well understood in research. But in practical application, the upstream and downstream of the academic pipeline remain disconnected.

From a societal perspective, if people do not understand how disasters occur and do not grasp a hazard’s “pathogenic mechanism”, they cannot effectively prevent and “treat” them. It is as though someone has researched the causes and treatments of “dental caries”, but no one does “public health” work to educate people and intervene in their health behaviours, and there are no “community doctors” to write prescriptions for individual patients – the “cavity” problem simply cannot be solved.

Recommended Report

Foodthink’s recently published report, Research on the Action Pathways of Rural Social Organisations Responding to Climate Change, previously detailed how the Zhuoming team responded to natural disasters and served as an “information hub”. Click the link to access and download the full report.

Foodthink Contributor
Pei Dan
A “word factory” worker who found her way back on track, focused on climate change, the natural environment, and the real people living through a world in flux.

 

 

 

 

Edited by: Ling Yu

Published by Foodthink