Exclusive: Unprecedented disasters are here, yet we are still ‘tasting a hundred herbs’

In the “humid zones” of the South, rivers with high flood risks generally don’t have such extreme variability—flows might be in the thousands or tens of thousands during the flood season and still be around two or three thousand normally, with banks flooding every few years. However, extraordinary floods with return periods of 20–40 years are much harder to prevent—human experience in flood control is difficult to pass down through generations.

The increase in extreme precipitation and over-standard floods is also changing how the government works. “Call-and-response” (alerting lower-level officials) has become a primary task for emergency management departments when dealing with rainstorms and floods. Once a red rainstorm warning is issued and flood response is activated, calls are made down the chain of command to the grassroots level: “(The) heavy rain is coming; get people in high-risk areas to evacuate immediately!” The first time this happens, it is inevitably chaotic.

With climate change, the 400mm isohyet is shifting north. In the future, in mountainous areas with stream and valley terrain between the 400mm and 800mm isohyets, the probability of extreme precipitation leading to severe flash floods and flooding of small-to-medium rivers will increase.

In August 2024, widespread rainstorms and floods in Jianchang County, Huludao City, and other areas led to at least 11 deaths. A team leader who participated in both last year’s Huludao floods in Liaoning and this year’s floods on the Beijing-Chengde border told me that in terms of type, scale, and on-site conditions, the two disasters were remarkably similar.

Can we learn from these disaster areas? Can we realise that it could happen to anyone, and then think about what needs to be done?

Hao Nan: Back in 2013, building on the experience of those before me, I proposed the “Five Imbalances“: imbalance in disaster types, imbalance between major and minor disasters, imbalance between urban and rural areas, imbalance across different post-disaster stages, and imbalance between pre- and post-disaster efforts. These five imbalances in social attention still exist today; nothing has fundamentally changed, except that whereas people once cared more about earthquakes, they are now more concerned with meteorological disasters. The allocation of resources has always been more closely linked to social visibility than to the actual scale of the disaster. Consequently, rural areas consistently receive less attention than cities.

Over the past decade, the trend of climate change leading to more frequent disasters has become obvious. However, the public only truly received a “lesson” during the extreme rainfall and flooding of 23 July 2021 in Henan (often referred to in the media as the “7.20 extreme rainstorms”). On the evening of 20 July, Zhengzhou’s urban area saw 200mm of rain in a single hour. This unprecedented deluge naturally caused severe urban waterlogging, and news of over a dozen citizens drowning in the metro triggered a public outcry. Yet, at the same time—or even slightly earlier—far more severe casualties occurred that few people knew about.

The official death toll for the Zhengzhou floods was 380, 70% of whom were from the hilly towns and counties in Zhengzhou’s suburbs, with deaths concentrated before 1 pm on 20 July. The vast majority perished due to flash floods, geological disasters, and flooding of small-to-medium rivers; only about a tenth died from urban waterlogging. Yet, public discourse focused almost entirely on the cases of urban flooding.

On the other side of the Yellow River in Xinxiang, the extremity of the rainfall around 20 July was no less severe than in Zhengzhou. Although the death toll was lower—as this is primarily a plain where water flows more slowly, allowing more opportunities for escape and rescue—the area was submerged in a vast sea of water. On satellite maps, villages became isolated islands, resembling Lake Poyang during its dry season. According to our preliminary estimates, including Hebi’s Qi and Jun counties where the floodwaters flowed downstream, nearly one million people were trapped by floods, with rescue teams assisting at least 400,000 of them. In the county-level city of Weihui in Xinxiang alone, 200,000 people had to be evacuated.

Hao Nan: After all these years, some critical issues remain undiscussed—the mechanisms and risk types of urban waterlogging are fundamentally different from those of flash floods, geological disasters, and river flooding. The risk of casualties from urban waterlogging is far lower than from flash floods or geological disasters, and the impact of the latter is usually far more profound. Even in Zhengzhou, despite such extreme precipitation, the urban floodwaters were drained in a day, while villages along the downstream rivers remained submerged for over half a month.

Furthermore, urban flooding mainly involves rainwater, which is relatively simple to clear. Urban infrastructure is typically restored within days, and citizens generally do not lose their livelihoods. In contrast, floodwaters rushing down from the mountains carry vast amounts of silt and debris, destroying everything in their path.

During the extreme floods in Qiandongnan Prefecture, Guizhou, this June, the silt mixed with rubbish left behind in Rongjiang city was one to two metres thick. More importantly, mountain floods destroy farmland, roads, and infrastructure. In the heavily hit townships of Xinglong County, Hebei, which we visited, many basic farmlands located in mountain valleys, as well as riverbank fields and terraced fruit trees on the slopes, were washed away. When basic farmland is destroyed by debris flows, it is difficult to reclaim and requires formal de-registration. While farmers receive some subsidies for losing land to disasters, the loss of livelihood is permanent.

In the future, the trend of frequent mountain floods will continue, and may even accelerate.

II. Promoting Disaster Reduction: Who can afford to spend three months drinking with the villagers?

“Disaster resilience” and “disaster prevention and reduction” are not the same conceptually, though this doesn’t mean they cannot be compatible in practice—at the very least, they can facilitate surface-level dialogue. As for donations towards disaster prevention and reduction, they have not increased over the last ten years; they remain very few, with projects being countable on one hand.

The return on investment for disaster prevention and reduction is difficult to measure, as there are few scientifically reliable benchmarks. Moreover, the more successful the effort, the more the situation after a disaster resembles a normal day, making the effect less visible. When benefits are not visible, no one invests. Public opinion easily spots losses that have already occurred, but losses that *could* have happened are rarely calculated seriously.

Ultimately, extreme disasters are low-probability events. If you implement disaster prevention and reduction in 1,000 low-risk communities, it is highly likely that no severe disaster will occur within the effective timeframe of those measures. To improve the effectiveness of investment in prevention and reduction, we must accurately identify high-risk communities—those with a higher probability of suffering severe disaster losses.

Disaster prevention and mitigation often only receive the attention they deserve after a disaster has occurred. However, it is very difficult to teach people how to handle a situation they have never encountered. And the unprecedented disasters that often cause the most severe losses are precisely those historic events that fall outside the local population’s cognition.

Experience in dealing with past disasters is crucial most of the time, but when facing an unprecedented disaster, it can sometimes become an obstacle. This year in the disaster-hit areas of Meizhou and Longyan in Guangdong, we saw that residents on both banks of the river used to move their ground-floor belongings and elderly relatives, who have limited mobility, to the second floor when floods came. But during the extreme floods last June, the water level rose above the second floor…

Only a handful of organisations with long-term rural projects can “incidentally” do disaster prevention. You can count such organisations on one hand, and they aren’t necessarily disaster experts. Some organisations just starting community disaster prevention projects ask, “Is there a PPT? Is there a course?” But this isn’t something you can just pick up and use.

Take the act of evacuation, for example; the goal is to move away from risk. But the nature of risk sources in a community—such as landslides, collapses, or heavy precipitation upstream—their attributes, dynamic mechanisms, and the subsequent impact area all require very specific professional expertise to judge. All of this influences the timing of evacuation, the choice of routes, and the formulation of contingency plans. Different types of disasters require different approaches, and the basic conditions of communities vary wildly. The various factors influencing the “awareness, belief, and action” of community residents are even more complex.

Many organisations are now starting to work on community climate adaptation, beginning with an assessment of the community’s meteorological disaster risks. However, everyone reports that this part is difficult to execute. From my observation, it is hard to integrate risk reduction or resilience improvement as a tangible project output.

Governments at all levels now place great importance on life safety. The foundational work for responding to meteorological disasters is being pushed hard, with responsibility passed down through every level. Departments for meteorology, water resources, emergency management, and natural resources are all requiring the grassroots to strictly implement disaster prevention and avoidance measures. For instance, according to the Ministry of Natural Resources, communities must implement “double control” for geological disaster “potential hazard points + risk zones,” and provide simplified guidance cards for disaster prevention and avoidance at hazard points. In theory, rural communities should also establish public welfare posts such as disaster information officers and geological disaster risk monitors.

Hao Nan: But implementation is equally difficult. During heavy rain and floods, mudslides and torrents at the source of the disaster often arrive at the same time as the warning, which is different from what is written in most emergency plans. Whether one can develop a method during normal times that actually solves the problem when a disaster strikes becomes the professional dividing line for community disaster prevention projects. Sometimes, community projects can serve to raise awareness, but the actual effect on disaster prevention and mitigation is harder to say.

An organisation once invited us to conduct a disaster risk assessment for a village. The risk background of that village was quite complex, requiring three or four vertical experts to carry out on-site surveys. For example, identifying the front and rear edges of a landslide—as cracks and heaving are precursors to a landslide—estimating the soil structure, understanding how thick and voluminous the potentially sliding soil layer is, which direction it will move, and which households it will affect. The budget for such an assessment was, at most, no more than 5,000 yuan.

Although the money was little, we were willing to participate in such a useful project, but we haven’t encountered another one since.

There is another type of community disaster capacity-building project that is part of a post-disaster support package, which is often a one-off. A friend recently went to Dingri County in Tibet for a community earthquake resistance project; there had been a 6.8 magnitude earthquake earlier that year, and project funding was available. These projects all follow a uniform pattern: they are only carried out where a disaster has already hit, and always within a year. It is essentially providing survivors with a sense of security and psychological comfort. Two years later, the locals no longer want to do it.

Disaster prevention projects, which should belong to the pre-disaster phase, can only be done after the disaster. Relying on the actual occurrence of a disaster to screen for high-risk communities is simply too costly.

III. Rocks Rolling Down the Mountain: Louder than the Shouts of Village Officials