METR 104:
Our Dynamic
(Lecture w/Lab)
(An Investigation):
Why Does West Coast Precipitation
Vary from Year to Year?

Dr. Dave Dempsey,
Dr. Oswaldo Garcia,
& Denise Balukas
Dept. of Geosciences
SFSU, Fall 2012

Part I: Analysis of Precipitation Records

The last two lab meetings of the semester, on Wednesdays, Dec. 5 and 12 (Lab Section #1) or Fridays, Dec. 7 and 14 (Lab Section #2), will be devoted to supporting the final project, which will be worth 15% of your final course grade. (Some of each of the last three lecture class meetings on the Mondays of Dec. 3, 10, and 17 will also support this assignment.)

The final project is a research project broken into four distinct parts:

  1. Part I: Analysis of Precipitation Data (described in lab during the week of Wednesday & Friday, Dec. 5 & 7).
  2. Part II: Statistical Connections between El Niño/La Niño Events and West Coast Precipitation (to be completed during the week of Wednesday and Friday, Dec. 12 & 14)
  3. Part III: Jet Stream Patterns during El Niño/La Niño Events (supported in lecture on Mon., Dec. 17)
  4. A final, summative report (due on Friday, Dec. 21; I will provide you with a template for this)

Overall Objectives:

Objectives for Part I:



Water is arguably the most vital natural resource for human and nonhuman life. This is particularly true in places where water is relatively scarce, such as much of the western U.S. The water that we depend on for drinking, irrigating crops, industry, and recreation, and that sustains the natural ecosystems on which we also depend, comes ultimately via precipitation.

The ways and places in which we live and work depend deeply not only on the average precipitation of the region but on its variability from month to month and year to year. To deal with variability and try to provide a reliable supply of water its residents, the state of California and the federal government have built a complex system of dams, pumps, and canals to capture, store, and divert water to places where people grow irrigated crops or prefer to work and live and where there wouldn't otherwise be enough water for those activities. However, California only stores in reservoirs or pumps out of the ground about half of all of the water that it currently uses, and it will be very hard to change that in the future. As a result, we depend heavily on the snow pack that accumulates in the Sierra Nevada Mountains (in eastern California) each winter to store water for us. As temperatures warm in late spring and summer, the winter snow pack largely melts, releasing the water gradually over a period of months, and we are able to capture and use some of it.

(Note that one of the likely impacts of global warming of greatest concern to California and other western states in particular, is on the winter snow pack. As the planet warms, more precipitation in the Sierra Nevada Mts. will fall as rain rather than as snow and will run off immediately instead of being stored as snow. We won't be able to capture as much of it as we need, and there will also be more winter flooding because the precipitation will run off in short periods instead of over several months.)

Most of the precipitation that falls in California, Oregon, and Washington is associated with midlatitude cyclones in the fall, winter, and spring. These storms form and travel along the polar front beneath the jet stream, so the location of the jet stream has a big impact on where midlatitude cyclones go and how strong they are, and hence where and how much precipitation falls. Although it's not possible to forecast individual storms with any confidence beyond a few days to (sometimes) a week in advance, there are some influences on weather patterns that are longer lasting and that can be predicted with some confidence months in advance. Sea surface temperature in the Pacific Ocean near the equator is one of those influences. Do sea surface temperature patterns affect the position and strength of the jet stream, and hence the path and strength of midlatitude cyclones, and hence precipitation patterns? And if so, how?

In this project, we are interested in the year to year variability of precipitation in California in particular but also other parts of the West Coast, and will try to determine what might account for some of the variability. There is reason to believe that the phenomenon of El Niño/La Niña, a quasi-periodic variation in equatorial Pacific sea surface temperatures that we can predict with some confidence months in advance, might West Coast rainfall. We will investigate the extent to which this might be true, and if it seems true, see if it might be connected to the position of the jet stream.

To investigate these connections, we will start in Part I by walking through an already performed, partial analysis of precipitation data recorded at several weather stations distributed along the West Coast.


  1. What weather stations will you use?

    You will use precipitation data from four weather stations, including one from each of the following four regions of the West Coast (see map):

    1. Southern California on or near the coast (San Diego, Los Angeles International Airport [KLAX], or Santa Barbara)
    2. Central California (Watsonville, Mission Dolores in San Francisco, or Sacramento)
    3. Northern California or southern Oregon (Eureka, CA; Ashland, OR; or Medford, OR)
    4. Washington state (Olympia, WA; Palmer, WA; or Bellingham, WA)

    We selected these stations because (1) they provide a representative distribution of stations up and down the west coast of the U.S.; and (2) each has a relatively continuous record of precipitation (missing no more than three days from any one month) from the current year going back to at least 1950. We got the data from the Western Regional Climate Center (

    Attached to this assignment is a list of particular station assignments for Lab Section #1 (Wednesdays) or Lab Section #2 (Fridays).

  2. Where can you get the partially analyzed precipitation data for your four stations?

    We've partially analyzed the data in Microsoft Excel (a spreadsheet calculating program) and saved them as PDF files. The files (one file per station) are accessible at

  3. How are the data organized?

    The instructor will illustrate how the data were analyzed in Microsoft Excel.

    Each precipitation data file contains monthly precipitation records for a particular station for each year from 1950 to the current year. The first column ("YEAR(S)") lists the years; the next 12 columns ("JAN", "FEB", etc.) lists the precipitation recorded for each of the 12 months; and the last column (labeled "ANN", for "annual total") lists the total rainfall for the whole year (all 12 months). Each row represents one year of observations. (Note that a number of months of the current year haven't been recorded yet, so those months are blank and no annual total is shown.

  4. How were the data analyzed?

    1. For each station, the average precipitation for each month from 1950 through the current time was calculated. The average annual precipitation for 1950 through last year was also calculated.

      (If you're interested in detailed instructions about how the analysis was done in Excel, see "Final Project, Part I: Excel Instructions".)

    2. For each station, a bar chart of the average monthly precipitation was plotted.

      (For the details about how this was done, see "Final Project, Part I: Excel Instructions".)

      How would you describe the variations in average monthly precipitation over the course of the year for these stations?

      What are the five months with the greatest precipitation at each station? If you had to pick five particular months that best identifies the "rainy season" for all four stations, what would those five months be?

      Note also that the five months that you identify won't necessarily be the five wettest months at all four stations, but they should represent reasonably well the rainy season at all four stations collectively.

      Note also that the there is nothing special about five months to identify the rainy season—in some places the rainy season might be longer and other places shorter, and in some places there might even be two "rainy seasons" (though probably not on the West Coast of the U.S.), but for the purposes of our research we want to standardize the definition because it makes comparisons less complicated.

    3. For each station, the total rainfall for each five month rainy season, from late 1950 through earlier this year, was calculated.

      (For details about how this was done, see "Final Project, Part I: Excel Instructions".)

    4. For each station:
      1. the data were sorted by rainy-season precipitation totals, from the highest to the lowest;
      2. the "wet" and "dry" years were color coded; and
      3. the data were re-sorted by year.

      "Sorting" the data means organizing it in some particular order. Initially the data were sorted by year, with the oldest data (1950) first and the most recent year last. We wanted to reorganize the data by rainy season precipitation total, with the rainy season with the highest total first and the season with the lowest total last.

      For the purpose of our research project, we'll define a "wet" year as any year in the top 1/3 of rainy season precipitation totals and a "dry" year as any year in the bottom 1/3. Sorting the data by rainy season total makes identifying the "wet" and "dry" years much easier.

      Color-coding the years means changing the background color of the cells in particular rows, so it's easier to tell them apart. The wet and dry years are easier to color code if they are grouped together, as they are after the data are sorted by rainy season total.

      For the purposes of using the Part I precipitation analysis for Part II of the Final Project, it is most convenient to have the (now color-coded) data re-sorted by year.

      (For details about how this was done, see "Final Project, Part I: Excel Instructions".)

      Your data are in a form designed to be analyzed further as easily as possible, in Part II of the Final Project.