30 March 2015

The atmospheric river that caused the Los Angeles flood of 1938

At least one person asked why I used a reanalysis that does not assimilate satellite water vapor data to study an atmospheric river (AR) event.

That's a good question because the NOAA/CIRES Twentieth Century Global Reanalysis Version 2c (20thCR V2c) only ingests three things: surface pressure, sea ice coverage and sea surface temperature. The rest of the analysis is generated by the physical models of NOAA's Global Forecast System (GFS).

The short answer is that 20thCR V2c extends all the way back to 1851, which means you can compare storm intensity between storms, including events that occurred before the modern satellite era.

E.g. how does the AR event of 2004-2005 compare with one responsible for the Los Angeles flood of 1938, that felled thousands of buildings?  Let's take a look.
Global animation of PWV from 20thCR V2c for 1938-02-27-00 UTC through 1938-03-03-18 UTC.
Again, I made a regional plot of the same time period, changing the color scale from [0.1, 70] to [0.1, 40] to better show the moisture gradients in the mid-latitudes.
Regional animation of PWV from 20thCR V2c for 1938-02-27-00 UTC through 1938-03-03-18 UTC.
In an era without satellites and numerical weather predictions, Angelenos had no advance warning. From KCET's Los Angeles Flood of 1938: Cementing the River's Future:
Between February 27th and March 3rd, 1938 Los Angeles was inundated with two storm systems delivering record breaking rainfalls. By March 3rd, the San Gabriel Mountains received 32 inches of rain, more than their average yearly total, and Los Angeles received over 10 inches of rainfall over the 5-day storm. 115 people lost their lives, thousands more were evacuated, over 6000 homes were damaged or destroyed, and 108,000 acres - one third of Los Angeles - was flooded.
Click to see more pictures of the storm damage.

The CSUN Digital Library offers more images, including this aerial photograph of Victory Boulevard in 1938.
Aerial view of 1938 flood damage in San Fernando Valley courtesy of CSUN Digital Library and used here with permission.
Compared to the visualizations of the 2004-2005 series of storms, the 1938 storm appears to have contained more moisture and over a wider area.

Of course, precipitable water vapor aloft is not a perfect predictor of localized rainfall.  That depends on terrain, lifting, and wind direction.  More accurate estimates can be obtained by ingesting reanalysis fields into numerical weather prediction models such as WRF.


  1. The connected historical sites are interesting, the approaches taken then would probably never happen today, e.g. straightening the LA river.

  2. This comment has been removed by the author.

  3. The long term interannual climatological impact of atmospheric rivers (AR) on the hydrology of California is also discussed in UC Berkeley professor B. Lynn Ingram's book "The West Without Water" in which she uses as a case study the Great Flood of 1862 (December 1861 - January 1862) when most of the Central Valley was completely inundated. Such events have prompted studies such as the U.S. Geological Survey's Multi Hazards Demonstration Project (MHDP) ARkStorm Scenario. Ingram also goes to great lengths to help the reader understand that the relatively "wet" conditions of the American west over the last 125 - 150 years are an anomaly (unfortunately encouraging westward expansion and rapid development) and that the long term interannual variability (100's to 1000's of years) is dominated by extensive droughts lasting decades to centuries. Perhaps the first person to realize this was John Wesley Powell, who strongly warned against rapid development of the west, especially for agricultural purposes dependent on abundant supplies of water.


This section is for people who want to discuss using our data holdings effectively. Moderators will delete irrelevant comments.