ArcGIS Living Atlas

Climate Extremes: Heavy Rains

Imagine what it would look like if everything you see in and around your house was suddenly buried under a foot of water. If you live in California, that is essentially what happened at the beginning of this month. From December 26th through mid-January, 11.5 inches of rain on average fell across the state. The heavy, sustained precipitation has already exceeded yearly wet-season averages, but the storms are also responsible for 20 deaths and heavy damage. More than an isolated incident, the latest climate models suggest that this trend of extreme, dangerous precipitation is the new norm. Let’s take a look at ArcGIS Living Atlas resources available to study immediate impacts and the long-term rainfall model predictions.

The Russian River floods in Northern California
The Russian River floods in Northern California

Short-term tools

The last of the heavy rainfall came through mid-January. ArcGIS Living Atlas tools can be used to understand these immediate impacts and as key decision support tools. Resources from the National Weather Service (NWS) in Living Atlas can bring to light these developing situations. For example, if you are reading this shortly after the January 2023 downpours, note the surplus of water in California rivers.

Living Atlas live feeds like the National Weather Service 72 Hour Precipitation Forecast shows accumulation totals in six-hour intervals. Below, see how the ninth and final atmospheric river battered Northern California over the weekend of January 15th.

Rain accumulates over a 72-hour period during mid-January 2023. Data is captured from the Living Atlas National Weather Service live feed. This period represents the ninth and last atmospheric river to hit the state in a three-week period.
Rain accumulates over a 72-hour period during mid-January 2023. Data is captured from the Living Atlas National Weather Service live feed. This period represents the ninth and last atmospheric river to hit the state in a three-week period.

Tools like these, providing near real-time weather forecast updates to a wide audience, will become increasingly relevant as our climate becomes more extreme. A recent Nature Climate Change article describes how intense West Coast winter storms could see a 40% increase in precipitation by mid-century. To investigate these trends yourself, you can look to historic data available in Living Atlas. The NOAA Storm Events Database 1950-2021 is the most comprehensive account of U.S. severe weather. Within ArcGIS Pro, you can filter event type to heavy rain and from there create a chart to visualize the trend by year. Since 1996, the number of heavy rainstorms has increased by about 44 events per year, which tracks with longer term trends.

Since 1996, heavy rain events have increased notably across the U.S. This data is derived from the Living Atlas' NOAA Storm Events Database. When filtered to "heavy rain" as event type, the data shows a trend of 44 additional heavy rain events per year since 1996.
Since 1996, heavy rain events have increased notably across the U.S. This data is derived from the Living Atlas' NOAA Storm Events Database. When filtered to "heavy rain" as event type, the data shows a trend of 44 additional heavy rain events per year since 1996.

Turning to future climate, ArcGIS Living Atlas has its own resources, with multi-model ensembles including multiple emissions scenarios and decade ranges, to help understand these trends. Here we will pull out some clear and interesting stories in the data and provide context for how to best interpret it.

More rain, nearly worldwide

Climate change means more heat. And while you may tend to associate heat with dryness (drought), more heat really means more evaporation, especially when that heat comes in the form of warmer oceans. Hotter air can hold more moisture, which means more rainfall. Think of this like the water cycle on steroids, dumping and evaporating more water at a faster rate.

Let’s take a look at some of the stories that emerge when we investigate the Living Atlas climate data. Once a subset of data is selected, you can perform raster algebra to look at absolute and relative change in variables compared to historic baselines. A previous blog explains in detail how to analyze multidimensional climate models, but the maps below serve as inspiration to see what stories emerge from the data. All analysis and layout was performed within ArcGIS Pro.

Annual changes

This hyped-up hydrologic system culminates with 93% of land surface getting more rain compared to historic baselines (1970-2000). The only notable precipitation decreases are in North Africa, the Amazon, and Central America. The greatest increases in annual precipitation occur in South Asia, a region that already sees heavy monsoon seasons, and (curiously) the Sahel region of Africa (a region that suffers from drought and desertification currently).

 

By mid-century, annual precipitation could increase significantly across most of the world. Looking at percent change against the historic baseline can show where future climate could differ most from typical, current patterns and is therefore a good indicator of the biggest shifts and greatest need for preparedness.
By mid-century, annual precipitation could increase significantly across most of the world. Looking at percent change against the historic baseline can show where future climate could differ most from typical, current patterns and is therefore a good indicator of the biggest shifts and greatest need for preparedness.

As explained in the heat component of this series on climate extremes, annual averages may conceal the potential impact of increasingly frequent, intense storms. In fact, most of the heavy precipitation by midcentury may not come in peaceful, quotidian showers but rather more dramatic, irregular events.

Seasonal variation

Though global climate models cannot show with precision when extreme events can occur, we can look at seasonal changes in precipitation to get a rough idea.

Breaking precipitation predictions into seasonal subsets shows that most increases in annual precipitation could be concentrated in the wet season. Nearly all of the world land surface could see increases in rainfall during their respective rainy seasons.
Even during the dry season, may parts of the world, especially the Northern Hemisphere, could experience increases in precipitation.
Breaking precipitation predictions into seasonal subsets shows that most increases in annual precipitation could be concentrated in the wet season. Nearly all of the world land surface could see increases in rainfall during their respective rainy seasons. Even during the dry season, may parts of the world, especially the Northern Hemisphere, could experience increases in precipitation.

In Living Atlas, bioclimatic variables 16 and 17 show precipitation by wet and dry season, respectively. The percent difference against a historical baseline for wet season shows that most of the overall increase in precipitation is concentrated during the wet season. However, even in the dry season, over 80% of land could actually see increases in precipitation.

Taking it LOCA(L)

In Living Atlas, downscaled LOCA climate model ensembles from NOAA provide not only US-specific data but even more granular variables for which to analyze precipitation extremes. Living Atlas has these variables for several emissions scenarios and years.

Annual Precipitation (USA)

To get us acquainted, let’s begin with the same annual precipitation variable, but this time calculated using 32-model ensemble tailored to the United States. This ensemble shows a similar trend compared as the global predictions. This data is reported by county, and the associated attribute table contains demographic information, allowing us to filter and perform calculations on affected populations. In deriving such statistics, you will find that 78% of Americans could see more rain by 2050 under the high emissions scenario.

Like global precipitation forecasts, the U.S.-specific LOCA climate data predicts increases in annual precipitation for most of the country.
Like global precipitation forecasts, the U.S.-specific LOCA climate data predicts increases in annual precipitation for most of the country.

The LOCA data really gets into the details, which is where the most acute climate impacts are found. The annual data tells us that Texas and Southern California may continue to dry out. Annually averaged, this may be true, but when you look at the annual number of days with high precipitation (2+ inches), a more nuanced story emerges…

Change in extreme precipitation frequency

LOCA data shows that the frequency of extreme storms could increase across the entirety of the lower 48 states by midcentury. Extreme storms are defined as those with more than 2 inches of accumulation. Even areas with predicted decreases in annual rainfall could see more frequent storms. Greatest increases in storm frequency can be found along the Pacific Northwest coast.
LOCA data shows that the frequency of extreme storms could increase across the entirety of the lower 48 states by midcentury. Extreme storms are defined as those with more than 2 inches of accumulation. Even areas with predicted decreases in annual rainfall could see more frequent storms. Greatest increases in storm frequency can be found along the Pacific Northwest coast.

The mid-century high emissions ensemble predicts that every part of the U.S. could see more extreme precipitation events. Even in areas with less rain overall (predicted decreased annual precipitation), storms could be more frequent. They may be as rare as one more every 8 years in the Great Plains, but an increase nonetheless. In other places (looking at you, Pacific Northwest) these heavy storms could happen more frequently (multiple additional storms possible every year).

Detailed view of extreme precipitation frequency increases along the California, Oregon, and Washington coastlines. A series of atmospheric rivers battered this area in early 2023.
Detailed view of extreme precipitation frequency increases along the California, Oregon, and Washington coastlines. A series of atmospheric rivers battered this area in early 2023.

The really interesting part is, if you zoom in to the West Coast, those California counties with more frequent heavy rain are the same counties that got absolutely inundated over the last month.

Change in consecutive wet days

Length of consecutive wet days may increase most along the Appalachian Mountain range. For this region, longer rainy stretches could be expected in the forecast. Most of the country, however, could see shorter consecutive wet days, which could mean more rain dumped in a short period (a dangerous combination).
Length of consecutive wet days may increase most along the Appalachian Mountain range. For this region, longer rainy stretches could be expected in the forecast. Most of the country, however, could see shorter consecutive wet days, which could mean more rain dumped in a short period (a dangerous combination).

Change in consecutive wet days is another neat variable that helps you parse out extreme risk beyond the annual and even seasonal averages. By county, LOCA ensembles show current and predicted length of precipitation in days. The county-level change in this variable indicates how much longer or shorter rain events may be. In most of the country, these events may actually be shorter, but in a few counties, mostly in lower Appalachia, there is a strong increase in storm length (in some cases days longer).

A closer look at watersheds

A more detailed look at subregional watersheds in the Southeastern U.S. Various levels of catchment areas are available in Living Atlas and provide context on rainfall accumulation and drainage.
A more detailed look at subregional watersheds in the Southeastern U.S. Various levels of catchment areas are available in Living Atlas and provide context on rainfall accumulation and drainage.

In all of these maps, we have included drainage watersheds also known as hydrologic units. Hydrologic units are nested features defined by hydrologic and topographic criteria are available in Living Atlas at the region, subregion, basin level, and so on. Hydrologic units provide useful information about flood risk since catchment shape influences how quickly storm runoff is accumulated in a watershed. Above, see a more detailed view of subregion watersheds.

Bivariate maps give us the ability to look at this risk intersection…

However, longer storms do not necessarily indicate more risk. In fact, it can be the shorter duration storms that dump more precipitation over a shorter period of time that can cause the most danger.  Therefore, the most risk exists at the intersection of more frequent precipitation and shorter duration storms. To figure out where this risk exists, we can invert the change in consecutive wet days (this time focusing on shorter periods) and combine that with change in precipitation frequency.

More intense, more frequent

The greatest flood risk exists when storms are frequent and dump a lot of rain in a short timeframe. This bivariate map looks at the intersection of high frequency and short duration storms to find the most compounded risk.
The greatest flood risk exists when storms are frequent and dump a lot of rain in a short timeframe. This bivariate map looks at the intersection of high frequency and short duration storms to find the most compounded risk.

This map shows where storms will be both shorter in duration and more intense. These heavy and increasingly common storms can cause destruction quickly, making flood preparedness essential.

Only two regions stand out for having more frequent and intense storms. The Mississippi Delta region where the river meets the Gulf is an area known for its destructive hurricanes and could see more intense rainfall. The second is the Pacific Northwest. Again, those recent Pacific atmospheric rivers may be just a sign of future weather trends to come.

Other Living Atlas resources for precipitation risk

When there is rain (and lots of it), there is flood. Therefore, any good storm preparedness is flood preparedness. The Flood Risk Index series within Living Atlas is a set of maps that show how different interventions could improve flood resilience. These layers focus on improvements to flood awareness, evacuation routes, open space, and impervious surfaces, among others. They tie in demographic data at the census tract level to provide socioeconomic context and information about disadvantaged communities.

 

 

About the authors

Abigail was a product engineer on the ArcGIS Living Atlas of the World team. Her research focused on the intersection of data science, climate change, and agriculture.

With over two decades of GIS experience Emily has mapped elephants in Thailand, wildlife poachers in the Republic of Palau, land use related issues around Yosemite National Park, and active wildfire incidents for the State of California. Presently she is a Senior Product Engineer and Cartographer with the ArcGIS Living Atlas of the World where she styles and designs layers, maps, and apps with the Environment Team. When not making maps, she is a true geographer and loves traveling with her family.

Connect:
0 Comments
Inline Feedbacks
View all comments

Next Article

Podcast 3- Hussein Nasser, Esri; When curiosity comes, give it a chance.

Read this article