Hong Kong is navigating a defining moment in its digital infrastructure development. The city’s data center market was valued at $3.62 billion in 2025 and is projected to reach $5.81 billion by 2031, driven by massive digital infrastructure expansion. As one of Asia’s leading connectivity hubs, Hong Kong continues to invest in the data center capacity needed to support growing demand for artificial intelligence, cloud services and digital infrastructure. This growth strengthens the city’s position as a regional digital hub but also raises questions about how it can meet rising demand while navigating climate volatility.
Rapid expansion also means greater exposure to climate risk. As Hong Kong builds the infrastructure needed to support AI and cloud computing, operators must contend with rising temperatures, more intense rainfall, stronger typhoons and heavy flash floods, all of which can threaten operational resilience and increase the cost of keeping facilities online.
Despite Hong Kong’s mature insurance market, investors still face a challenge. Historical data shows how assets have performed in the past, but it cannot fully account for how climate risks are changing. As data center investment grows, forward-looking climate analysis will play a bigger role in assessing long-term resilience and investment risk.
Climate change is expected to intensify extreme rainfall and raise sea levels, increasing the likelihood that flood events will disrupt critical infrastructure and essential services. In Hong Kong, this risk is particularly evident because flooding is increasingly driven by highly localized urban flash floods. The city’s steep, mountainous terrain causes intense rainfall to run rapidly into heavily developed districts, creating fast-moving flood events that can overwhelm drainage infrastructure. These highly isolated events are also among the most difficult to model accurately, as relatively small factors, including blocked drains or concentrated bursts of rainfall, can significantly alter flood behavior.
Data centers, which must run 24/7 to service data needs and justify cost, are frequently designed and constructed in areas based on a patchy, inconsistent field of local flood hazard data. This is compounded by these data often being out of date and existing within a “nonstationary” system, owing to the impact of global warming.
Moreover, there are predictions of a “super” El Nino cycle where ocean temperatures rise two degrees or more above historical averages. In Hong Kong, despite seasonal rainfall totals likely being reduced, the resulting warmer atmospheric conditions are expected to make episodes of extreme localized rainfall more intense. This occurs due to a fundamental thermodynamic law known as the Clausius-Clapeyron relationship: A warmer atmosphere holds roughly 7 percent more moisture per degree of heating. This increased water-holding capacity acts as a planetary sponge, squeezing out intense, concentrated deluges that easily overwhelm conventional urban drainage networks. When the eventual transition to La Nina occurs, Hong Kong is more likely to experience an increase in cyclones during the peak season, heightening the risk of flooding and disruption to critical infrastructure.
For a standard commercial asset, you can frequently engineer your way out of a localized flood risk with raised foundations or concrete defenses. However, for a data center, the operational liabilities extend far beyond the property footprint. Resilience depends on the infrastructure that keeps the facility operating, including power, cooling systems, substations, telecommunications and fiber networks. In Hong Kong, where flash flooding can disrupt infrastructure well beyond an individual site, these interconnected assets can become the first point of failure. If the local energy grid or underground fiber-optic network is damaged by flooding, the impact extends far beyond the building itself, creating costly business interruption even where the facility remains operational.
Historical flood maps and data remain valuable, but they are becoming a less reliable guide to future flood risk as climate conditions change. In Hong Kong, where highly isolated flash flooding is difficult to model, forward-looking probabilistic analysis is becoming important when assessing long-term infrastructure resilience. Flood models are data-hungry things, and they must be built with highly precise information.
Hyperscale tech leaders like Microsoft have already realized this, bypassing legacy maps entirely to utilize advanced probabilistic modeling to safeguard their global server assets. To bridge this protection gap, the regional infrastructure sector must upgrade its corporate view of risk by transitioning toward forward-looking probabilistic modeling which utilizes machine learning algorithms, combined with physics-based models, to provide flood risk data that is truly fit-for-purpose.
Historically, major infrastructure players assumed that sheer balance-sheet capacity and over-engineered physical defenses could insulate an asset class from localized perils. To survive this impending climate cycle, regional leaders must confront the uncomfortable truth that this assumption is no longer the case.
While this volatility will not stop Hong Kong’s data center boom, it does have the potential to reprice it. Business interruption and infrastructure damage could drive up insurance premiums and operational costs, translating directly into a higher “cost per gigabyte”. At the same time, investors, lenders and insurers increasingly expect organizations to understand and quantify their exposure to physical climate risk, while financial regulators are placing greater scrutiny on how businesses measure and stress-test climate-related risks.
As the Earth’s climate system continues to reach unprecedented levels of warming, we cannot rely solely on historical records to assess risk. For Hong Kong’s data center sector, long-term resilience will depend on forward-looking climate assessments that help operators understand how future flooding patterns could affect both facilities and the critical infrastructure they depend on.
The author is a hydrologist, cofounder and chief operating officer of Fathom, a Swiss Re company.
The views do not necessarily reflect those of China Daily.
