Old Route 49 bridge crossing over the South Yuba River in Nevada City, Calif. saw local and regional visitors during the atmospheric river event across Northern California on January 9, 2017. 

What is a “megastorm?” For the western USA that means an uncommon storm system that delivers extremely high levels of precipitation – mostly rain – over an unusually long period of time, over a broad regional area, causing “truly massive floods” as scientist B. Lynn Ingram put it. In addition, a megastorm may produce hurricane-force winds with more speed and power than almost any other ‘western’ storm system. It may also produce coastal storm surges much like a hurricane does, compounded by the amount of excess water that is flowing from land into the rising sea.

The term ‘Atmospheric River Storm’ is sometimes used, but Atmospheric Rivers (“ARs”) can actually be relatively small or large events. UCLA researchers including D. L. Swain have called a megastorm a “severe storm sequence” to distinguish them from “extreme wet years,” which are smaller and about eight times more common. Another term, “ARkStorm,” explained below, has been popularized in recent years, but strictly speaking that name refers to one particular computer simulation that was extensively studied for planning purposes by the United States Geological Survey (USGS) in 2011. California experienced its most recent megastorm, lasting 43-45 days, in 1861-62.

Atmospheric river systems were only discovered about 20 years ago. Today’s satellites have sensors that can study the Earth’s atmosphere in detail. They have found bands of very moist air that develop around the Earth in the tropics and can move irregularly toward the poles, about 1 mile above sea level. Water-saturated air sometimes peels away from the tropical Western or Central Pacific and bends northeasterly across the ocean to strike our West Coast, typically in California.

In the past two decades, movements of those air currents have been observed and analyzed using satellite data, and are now tracked in real time and used in 7- and 10-day weather forecasts. Smaller atmospheric river flows are fairly common and provide a significant portion of the precipitation in the western USA. Big atmospheric river flows are less common but can be dangerous: “AR” systems can bring water onto the West Coast at rates as high as 1 to 15 times the average rate of water flowing out of the Mississippi River (which drains more than 1,245,000 square miles of the USA!). The weathercasters’ cute term “Pineapple Express” trivializes a potentially powerful weather system that can sometimes have catastrophic consequences.

In 2011 the USGS released a report on a study project involving more than a hundred scientists and experts, trying to understand these megastorm patterns and their impacts on modern California. Basically, the USGS team tried to simulate conditions roughly similar to the known megastorm of 1861-62. Weather records from the Gold Rush era are limited, so the USGS-led scientists basically combined good data from two actual “extreme wet” storms that hit California in 1969 and 1986, together totaling 23 days of rain (versus the actual 45 days in 1861-62), which might roughly match the incomplete rainfall reports from 1861-62 (the 2011 computer simulation described an apparently smaller megastorm that still – if real – would be extremely dangerous). Later in this series I’ll explore the many serious storm impacts predicted by that study.

For their storm scenario, the USGS team coined the term “ARkStorm,” which deserves a bit of explanation. “AR” stands for Atmospheric River, while “k” represents 1,000 (like “kilo”), an arbitrary number to indicate size (something like the Richter Scale for earthquakes). For example, if another smaller or larger storm, real or computer-modeled, were compared to the 2011 ARkStorm simulation, it might be scored as a 900 or a 1,280 when compared to the 1,000 of the ARkStorm baseline. It does not mean a “1,000-year storm” (some writers have confused this point, unfortunately). In fact, calling things “thousand-year storms” causes nothing but trouble in our thinking, as I’ll explain here.

In 1968 Congress and FEMA launched the National Flood Insurance Program (NFIP) which, using statistics, prepares flood likelihood estimates and maps to help lenders and insurers (including Federal programs) profitably or responsibly manage financial risks when insuring property that might flood. River flow and flooding estimation data and mapping are also used for planning community development and flood controls, and for forecasting “tolerable” stresses to infrastructure such as dams.

To provide “the odds,” the agency uses a sample of actual stream flow records from river gauges. These are used to plot a curve of peak rainfall and snowmelt stream flow patterns during the “sample” years. A computer then calculates a prediction from the curve for future years, to estimate the maximum amount of water that might be predicted to flow in any one year out of 100 (not “every hundred years”) or 500; mapping can plot where that water is likely to go. But these predictions can be affected by how the sample and the prediction curve are chosen. The ARkStorm study pulls back the curtain on the problem.

For their comparisons, the USGS ARkStorm study used this same estimating method on an 87-year sample set of river data from 1916 to 2003. They found that in a megastorm like the ARkStorm, central Sierra rivers would flood at levels up to about what standard methods would predict for “1 year in 1000.” Yet, as I’ll show next week, megastorms are far more common than that. The sampling problem here is that no western USA set of river or rainfall data, like 1916-2003, contains a year like 1862 that had a megastorm.

Twentieth-century data curves are missing megastorms. The problem is like measuring local windspeeds every day for five years, then designing buildings without any concept of tornados or hurricanes. The 1-in-100 risk estimates based on “typical” weather don’t apply to all situations.

About those odds: Recently there have been a number of studies on the “history” of California weather, supplementing the oral histories of the Mi-Wuk people, and establishing the hard evidence for dating other megastorms prior to 1861-62. Next week I’ll summarize important new predictions from those data. Spoiler: megastorms aren’t all that rare.

In providing this information, the author’s intention is to motivate and support awareness, discussion, and preparation, not to cause anxiety or distress. Knowledge, planning and preparation can overcome many difficulties. It may help to think of possible future emergencies as “What if…” scenarios for problem solving. “If the 1861-62 megastorm was not the only one of its kind, what else to I need to know before I can wrap my head around this stuff?”

Be sure to check out Part 3 of Rick Kerrigan’s series on California Megastorms - Are megastorms flukes? If not, how frequent are they?– in next week’s Ledger Dispatch. For links to sources and further information, see below:

2011 USGS ARkStorm Study report:

Atmospheric River videos:

National Flood Insurance Program



Floodplain development: