LETTERS

The Decade of 1,000-Year Floods

April 6, 2023
The Decade of 1,000-Year Floods
The Decade of 1,000-Year Floods

California is emerging from an extraordinary winter. Many parts of its eastern mountain range, the Sierra Nevada, have seen exceptional snowfalls and the deepest snowpack on record in at least 90 years.  At lower elevations, equally exceptional rainfalls have refilled parched reservoirs and greened a landscape that is often brown.  A year ago, 100% of the state was experiencing drought and more than 40% of the state was experiencing extreme drought; today, less than 2% of the state is experiencing dry conditions, in its naturally driest spots. The state’s near-term water fortunes have changed dramatically in the course of just four months of winter storms. 

The state is still in a mega-drought, characterized as a decades-long period of low precipitation and low soil moisture, but at the moment, it is well-watered.  

While this winter is extraordinary, it is not unprecedented in the historical record. Nor, for that matter, is it unexpected in the future, in California or elsewhere.  Devastating floods coursed through Europe in 2021. In 2022, flooding in Pakistan destroyed or damaged two million homes, 13,000 kilometers of highways, 439 bridges, and more than 4 million acres of agricultural land; the value of those losses equal 2.2% of the country’s GDP. The world is warmer than it was a century ago; a warmer planet holds relatively more atmospheric moisture while also being relatively drier at the earth’s surface. The result is that once-exceptional precipitation events are becoming more frequent.  We have entered a decade of 1,000-year floods. 

California is the world’s sixth-largest economy, and it only exists thanks to almost two centuries of intensive management of water. It is a case study for the extreme water concerns that other parts of the world will soon face, if they are not experiencing them already. This letter examines what this decade – and decades to come – mean from a broad climate lens, including climate adaptation and mitigation, impacts on power and infrastructure, implications for our global food and agriculture system, and the role that technology can play in preparing us for this new world.   


I. The Atmospheric River

The phenomenon creating California’s precipitation pattern is called an atmospheric river – a long, narrow region of the atmosphere carrying moisture outside of the tropics. The name is apt; an average atmospheric river carries as much moisture as the Mississippi River at its mouth, and an intense atmospheric river carries as much as 15 times that much moisture. Because atmospheric rivers are relatively narrow in reach, the moisture they carry tends to be concentrated as it falls. 

California has now experienced more than a dozen atmospheric rivers since the start of this winter. While much of this precipitation fell as rain, it has fortunately fallen as snow at higher elevations, where it will remain relatively locked away for months to come. Meteorologists classify the first of April as the end of California’s snow season, and as of that date, the state’s snowpack was more than 230% of its normal levels. Statewide, California’s snowpack is at a record-high level.

Chart, line chartDescription automatically generated

With much of that precipitation falling as rain, California’s reservoirs have also re-filled substantially, a great relief after a very arid summer last year. As of April 1, 12 out of the state’s 17 major water supply reservoirs were above their historical average levels for the date. If more rains arrive – certainly as the Sierra snowpack melts – these reservoirs will fill.  The state may then be presented with the challenge not of too little water in reservoir storage, but too much. 

In the past twelve months, California’s Central Valley, home to not only much of the state’s agricultural activity but much of the nation’s, has emerged from an extreme drought.

ChartDescription automatically generated with low confidence

This dramatic transformation is clearly visible from space. In August of 2022, the state’s central areas were green only where irrigated; at the end of March, almost the entire Central Valley was lush, and the Sierra Nevada snowpack is quite evident. 

The immediate reaction for much of the state is relief. Rains have fallen onto parched soil and snow has fallen onto cold mountains now holding 5.5 feet (1.37 meters) of stored water. But the future is more complex – and to see it, we must look to the past.  


II. This Lake Does Not Exist

One of Voyager’s favorite media artifacts is an image tweeted by the author Douglas Coupland nearly a decade ago. Coupland entitled it simply “California 1850”, and it shows an imagined well-watered state without any of the agricultural and infrastructure work that would later transform one of the nation’s most distinctive geographical features: the Central Valley. The image, created by photographer Mark Clark, is a vision of the state a few years after the Gold Rush, and on the cusp of profound transformation.

A picture containing natureDescription automatically generated

Source: Mark Clark, via Big Think

The Central Valley of California is 400 miles long and 50 miles wide (650 kilometers and 80 kilometers, respectively), bounded by mountains in every direction. It is very flat, and prior to intensive human habitation and agriculture, was a mix of prairie, desert grassland, forest, and extensive marshes and lakes. One of those lakes appears like a phantom in Coupland’s map – so much a phantom that when Voyager first encountered it, we said “this lake does not exist”. 

But it does, in fact. It is Tulare Lake, once the largest body of water west of the Mississippi by surface area. Fed by seasonal melt from the Sierra Nevada mountain range, it was four times larger than Lake Tahoe far to the north (in surface area, if not volume).

The lake was not just a lake; it was part of a system of connected water bodies that ran from the southern end of the Central Valley into the San Francisco Bay. When José María Narváez mapped California in 1830, he described the Central Valley as “Ciénegas Tulareus” – ciénega being not a lake, but rather a swamp or slough. Two decades later, when explorer John C. Frémont mapped the same area, he described only the lake, still massive. 

By the end of the 19th century, human activity began to erase the lake. Water diversion projects sent its water sources elsewhere, and landowners began to carve away plots of land even as they were still within the lake’s geological confines.  An 1876 survey map of Tulare County shows lots extending into recently filled areas and even the lake itself. A century and a half ago, Tulare Lake was disappearing from the map.

1876 survey map of Tulare County, California

MapDescription automatically generated


Source: David Rumsey Map Collection link

On its way out, but not permanently. Already in 2023, 30 square miles (78 square kilometers) of Tulare Lake has re-emerged. That emergence is before the all-time record southern Sierra Nevada snowpack has begun to melt, forcing water into the southern Central Valley without any outlet except the California Delta hundreds of miles to the north. Locals are not being hyperbolic when referring to the snowpack to their east as a ‘ticking time bomb’. With the melt, Tulare Lake could reach seven times that size later this year. Humans may have drained the lake, but a warmer atmosphere will refill it. And continued warmth, combined with centuries-old weather patterns, could make the lake into something even more massive. 


III. The Mega-storm

California’s historical record is clear: every few hundred years, the state experiences extreme precipitation events with the potential to re-write the physical landscape. 

More than a decade ago, University of California geographers used radiocarbon dating of a lake in the Sacramento River floodplain to determine that enormous floods took place from the years 1235–1360, 1295–1410, 1555–1615, 1750–70 and 1810–20. After that, there was one more storm, one that intersected with the start of the state’s major population and infrastructure expansion, in 1861-62. 

That storm is oddly little-remembered in the United States, perhaps because it happened at the western periphery, and perhaps because it coincided with the start of the Civil War thousands of miles to the east. That storm flooded the entire Central Valley, destroyed a quarter of the state’s economy, and forced California into bankruptcy. California’s capital, Sacramento, was flooded for three months. Paleoclimatologist Lynn B. Ingram noted that a storm late in 1861 featured 10-15 feet (3-4.5 meters) of mountain snow followed by intense, warm rain. The result: 

As floodwater gathered in the valley, it formed a vast, muddy, wind-roiled lake, its size “rivaling that of Lake Superior,” covering the entire Central Valley floor, from the southern slopes of the Cascade Mountains near the Oregon border to the Tehachapis, south of Bakersfield, with depths in some places exceeding 15 feet.

This storm not only altered the physical geography of the state – it also changed the human geography of the Central Valley.  Prior to the storm, much of the land was still owned by Mexican rancheros, landowners whose land titles dated back to the Spanish rule of the area. The rancheros had done well during the Gold Rush, supplying beef to mining towns, but had suffered during droughts starting in 1856. The flood, according to historian Lawrence James Jelinek, forced them off their land and also forced them to sell to white settlers “for pennies on the acre”. 

As settlers displaced rancheros, they also changed how they used the land. The water that destroyed the valley during mega-storms also provided its horticultural life in gentler seasons, which farmers used to convert prairie and marsh into planted fields. The Central Valley’s current agriculture was born of flood, and is maintained today by irrigation and intensive water management. 

If California’s historical weather patterns provide the fuel for another mega-storm, then climate change provides the accelerant. As a 2022 paper by Xingying Huang and Daniel L. Swain notes, climate change is increasing the risks of a mega-flood of historical magnitude in California. The researchers state, 

We find that climate change has already doubled the likelihood of an event capable of producing catastrophic flooding, but larger future increases are likely due to continued warming. We further find that runoff in the future extreme storm scenario is 200 to 400% greater than historical values in the Sierra Nevada because of increased precipitation rates and decreased snow fraction. These findings have direct implications for flood and emergency management, as well as broader implications for hazard mitigation and climate adaptation activities.

And thanks to greater economic value and greater concentration of impacted infrastructure in contemporary California, a mega-storm could be extremely costly. A 2019 study estimated the cost of a new mega-storm in California at $725 billion, three times more than the cost of a severe San Andreas Fault earthquake and five times more costly than Hurricane Katrina in 2005.   

California’s built environment has always been subject to extreme weather, which in the historical record is not only relatively frequent, but also overdue given historical patterns. Climate change has given that extreme weather more atmospheric energy, and human activity has created a much greater swath of economic activity at its mercy.

We can plan for this mega-flood future. We must. 



IV. Building for a Warmer, Drier, Wetter, Denser Future

Preparing for the decade of 1,000-year floods requires strategy, planning, and investment in many dimensions. It will require greater awareness of how a warming climate interacts with our earth system, faster and more accurate measurement of changes, longer-duration predictions of weather patterns and their impacts, a hardening of infrastructure, a concerted response to disaster, and above all a willingness to explore new ways of doing established human activities. 

Voyager thinks of these changes, and responses, as most relevant in the following areas. 


Preparing for the storm 

The world will need ever-better ability to predict severe weather patterns and anticipate their impacts. Weather prediction has improved greatly in the past few decades – fortunately, the atmospheric river activity in California this past winter was mostly seen developing in real time.

However, understanding of local impacts can always be improved. That means greater awareness not just of where a storm might land, but also the characteristics of the land on which it is falling. For built-up areas, that means more precise mapping of potential flood impacts. Today, there are more than 14 million properties in the US alone are at risk of annual flooding, regardless of what current (and outdated) government flood maps say. That planning begins with better data, such as that provided by companies like DeltaTerra and Floodbase, which are filling the void left by antiquated federal flood insurance programs.   

At ground level, better understanding means a recognition of local soil characteristics, to start. It would also include much more precise mapping of potential flood impacts, given that many flood maps are decades out of date. Relatively mappable information such as soil moisture content will help determine how much water an area can hold and the speed with which it absorbs it. There will also be opportunity for sensing ocean, and not just atmospheric, patterns. Given that California’s atmospheric rivers originate close to Hawaii, knowing ocean conditions thousands of miles away will be very useful.

We will need to develop a much better understanding of how other aspects of our built environment or agricultural land might respond to inundation. That means knowing the impacts on the water table and aquifers if land is inundated for days or weeks, but also what standing water will interact with during that time. Heavy metals, agricultural and industrial waste, and high salinity are all found in cities and in heavily planted- and farmed soil. When they contaminate surface water, the consequences can be disastrous – and lasting – for human, animal, and ecological health.  The first step in preparation is greater knowledge. 

Finally, policy at all levels must acknowledge the real and rising risks to people and their property. Sea level rise is inevitable (and already evident in US locations such as south Florida, both in sunny-day flooding and in the insurance premia paid for vulnerable buildings). There is no wishing it away with climate denialism, however convenient the politics may be. 

California’s Central Valley has a large population without the resources to self-insure against catastrophic loss. Recent research from the US Federal Reserve, Environmental Defense Fund and First Street Foundation finds that climate risk isn’t being priced into the US housing market, with disproportionate potential impact on low-income households. Additional research suggests that a full 20% of US homes have “meaningful exposure” to mispricing from flood risk. Beyond planning and permitting to reflect real flood risk, financial instruments will need to include more than just insurance products.  We will need political alignment, better risk data, and innovative debt instruments to help communities respond in advance. 




Reacting to the flood

Reacting to the rising water – whether from an immediate storm or from months of melt – will require detailed planning and preparation. 

Planning starts with understanding how impacts will unfold over time. Rains fall today, but snow might not melt for months, and when it does, it may not do so evenly. Greater resolution on snowpacks, how they melt, and when and where the now-melted water will flow will be essential. 

Some of this capability will be a matter of better sensing. Today, that is accomplished by human overflight in sensor-laden planes. This is useful, certainly, but does not provide a constant stream of data. In the near future, drone-based sensing could potentially gather this information with greater frequency and with less risk to human operators. Insights could be extracted from this information with higher sophistication through more advanced modeling of both long-term and acute impacts. 

Preparation means building readying the environment where possible. That can look as simple as sandbags around a house, or the shoring up of levees to channel floodwaters. It can mean long-term changes to infrastructure, such as planning building stock not just around the 100-year flood, but the 1,000-year flood. It means preparing supply chains for potential disruptions as far in advance as possible, by diversifying supply and increasing an understanding of agricultural stores available to meet flood-induced shortages.  And it will certainly mean better financial instruments to price flood risk. 

In even the wealthiest regions worldwide, very little of the economic loss from major flood events is insured. In last year’s floods in Pakistan, none of the $14.9 billion of economic losses was insured, and in Europe’s 2021 floods, only 25% of total losses were insured. 

Chart, bar chartDescription automatically generated

The practice and policy of water management

Ideally, the same water that imperils infrastructure during floods can become a resource for re-charging not just reservoirs, but the water table as well. Enabling that will be a matter of better infrastructure, but it should also include distributed technology, coupled with thoughtful planning. That means planning for permeability – enabling our built environment to pass more water into the soil and thence to the water table, rather than rushing it away. That is a physical opportunity (for new surface materials that are both permeable and durable) and a permitting opportunity (to allow cities and private developers to incorporate new materials). It is also a nature-based opportunity, including restoration of wetlands that can slowly absorb the flow of floodwaters while also filtering it. 

There is a substantial energy element to water management as well. Moving water across California water may account for as much as 10% of the state’s greenhouse gas emissions. As of 2018, 20% of statewide electricity demand and 30% of statewide business and home natural gas demand goes to pumping, heating, and treating water. The future requires ever-greater efficiency in moving water across long distances, treating it, distributing it locally, and decarbonizing the energy needed for each step.  Voyager may examine the energy-water connection in a future letter. 

Addressing extreme precipitation as a potential resource must be done in the context of water rights, resource rights, and ecosystem sustainability. Doing so requires planning across precipitation cycles. This year’s West Coast ban on chinook salmon fishing is an instructive example: due to last year’s extremely dry conditions, the forecast for fish returning to rivers to spawn is near a record low. Next year should be better, thanks to this year’s river flows – all the more reason to have durable but supple plans that can incorporate community needs over time. 



Climate mitigation meets climate adaptation 

A warmer climate is not just an input to the decade of 1,000-year floods.  Floods themselves have the potential to impact climate change and our responses to it. 

Extreme precipitation will have direct impacts on climate change.  Areas that are submerged and lead to rotting organic matter will produce methane, a highly potent greenhouse gas. Bursts of vegetation during wet years create the fuel for forest fires in hot, dry years. Just one year worth of fires in California, in 2020, probably negated the climate benefit of decades of reduced emissions from fossil fuel consumption. Better sensing of risk areas after floods, and predictions of potential fire spots, will be necessary to manage precipitation-induced greenhouse gases. 

Chart, bar chart, histogramDescription automatically generated

In addition, today’s energy-related climate mitigation investments are at risk from extreme precipitation events. Power equipment responds poorly to prolonged inundation, and it is not just California that has many gigawatts of solar generation assets sited in historical floodplains. Preparing for floods can be done – but it involves building hardened infrastructure and elevating mounting structures far above ground, which incurs not just additional construction costs but additional operations and maintenance costs as well. Offsetting those costs is a potential opportunity for new construction techniques and new automated O&M. 

In essence, preparing for 1,000-year floods will mean melding climate adaptation to our instruments of climate mitigation. We will need to build a decarbonization system that is hardened against the impacts of today’s already-warm climate, and what it will bring us tomorrow. 



Planting and harvesting 

California’s Central Valley is a special agricultural region.  While it is just 1% of US farmland by area, it is 17% of the nation’s irrigated land. It provides 8% of the nation’s agricultural output by value, a quarter of its food, and 40% of its fruits and nuts. It is also 12.8% of the US agricultural exports by value, including 100% of small exports like pistachios, artichokes and olive oil. It is a significant exporter of major crops too: the Central Valley exports almost a third of US dairy and dairy products by value, and 40% of US rice. California on the whole exports 90% of US wine by value, though most of that comes from other regions of the state. A mega-storm could at the very least impair a massive amount of domestic food production – as Pakistan’s floods did last year. 

The agricultural response to major floods will need to be multi-pronged. It requires upgrading infrastructure to adapt to potential periods of extreme precipitation or inundation. That infrastructure would not just be earthworks to prevent flooding – it could include warehousing and storage, or nimble supply chains to quickly move products from sources to demand centers. 

It could also require growing food differently. That could mean scheduling planting around expected wet and dry periods, where long-range forecasting is available. In the longer run, it may mean breeding resilience into agricultural crops – creating crops that can withstand both prolonged dry and wet periods. 



V. One Decade, A Thousand Years

The decade of thousand-year floods is here. It is the result of natural patterns, amplified by anthropogenic climate change, and multiplied an ever-larger built environment and agricultural system.  

Extreme weather is inevitable as long as global greenhouse gas emissions stay at current levels and atmospheric GHG levels remain at their highest level in at least 800,000 years.  Extreme loss and damage, however, is not inevitable. Responding to the decade of 1,000-year floods will require concerted effort across multiple sectors. It will require foresight, creativity, invention, and re-imagination of how and where we build, grow, and harvest. 

Voyager believes that meeting that challenge is an essential response to a wetter and more volatile world. It is also an opportunity to reimagine how human activity interacts with the natural world. It will require adapting to volatility in the near term, and minimizing it in the long term, creating resilience and ensuring stability as yesterday’s exceptional floods become tomorrow’s normal.