Advancing Water Resources Research and Management

Symposium on Water Resources and the World Wide Web
Seattle, Washington, December 5-9, 1999

Climate Analysis Resources From the Climate Diagnostics Center

An important part of environmental resource management is the ability to put current climate data (such as seasonal precipitation) and climate forecasts into a historical perspective. This perspective can be in terms of the climatology, the annual cycle, and/or various known statistical relationships (for example ENSO's effect on precipitation). At the NOAA-CIRES Climate Diagnostics Center (CDC), we have available a variety of web resources that allow users to interactively monitor and assess the current climate and allow the user an opportunity to compare that climate to that of the historical record.

CDC's climate resources can be divided into four parts:: Climate Monitoring
Climate Data
Climate Forecasts
Analyses of Historical Relationships

Climate Monitoring

At CDC, we monitor the current climate with many different products. On an ongoing basis, we maintain a "weather map room" and a "climate map room". In the first, we provide maps of various weather variables and satellite pictures. We also provide short-range (0-4 day) and long-range (up to 2 weeks) forecasts. The climate map room products monitor the climate on longer time-scale (7 days to seasons). There are also items of a more technical nature (for example recent storm track locations) that reflect research in progress at CDC.

In addition to our more general weather and climate pages, we monitor current important climate events. Our web pages describing the existing La Niña are an example of this type of activity. Other climate events which we have showcased included the 1998 Texas drought and the 1999 summer east coast drought. The latter two have evolved into a general precipitation monitoring page whose purpose is to keep track of drought/flood events in the US over various time-scales.

Climate Data

CDC distributes various climate data-sets at no-cost via the Web. Our largest data-set is the NCEP/NCAR Reanalysis, which consists of gridded analyses of atmospheric variables at multiple atmospheric levels 4-times daily from 1948 to the present. These fields are useful for more closely examining many processes including determining dynamical relationships, budgets (e.g. water budgets), and synoptic events. They are also useful for looking at regions with no or little measured data. Other data-sets that are in our archive include the COADS long-term ocean data-set, operational model output, various SST and rainfall data-sets and Outgoing Long-wave Radiation (OLR). Some of these data-sets can be down-loaded via FTP while others must be ordered for delivery on tapes.

All of our data-sets are in netCDF, a self-describing format that is supported on a wide variety of computer platforms. The self-describing feature of this format makes it relatively straightforward to create web-based tools that can handle all of our data, without writing a lot of data-set specific code. An example of such a tool is our Web atlas interface which allows users to look at selected fields of our data before down-loading the data files. This interface was used to create the plotted subsection of the CPC Merged Analysis of Precipitation (CMAP) data shown left.

In addition to a quick look page for individual data-sets, a detailed search page is available that enables the user to search for data by data-set and variable. The results of this search guide the user to the web atlas interface so they can plot the data or down-load `

Climate Forecasts

Plot of long-ranage precipitation forecast chance

CDC provides probability forecasts based on the concept of multiple model runs ("ensemble" runs). Instead of a single forecast, multiple runs are done using slightly different initial conditions. These different initial conditions are used to represent unresolved initial conditions due to spatial sampling and grid sizes. By looking at the collection of runs, the user can get a more accurate idea of the of chance of an specific occurrence than they can by looking at the output of a single model run. The example at left shows a typical climate forecast. Specifically, it shows the odds of the temperature being above or below normal temperature based on multiple runs of the climate model.

In a direct application of our research results, we provide a tropical Pacific and Atlantic SST forecast. These forecasts, which are based on statistical methods, have proven comparable to results obtained from numerical models.

Historical Relationships

Plot of correlation of winter tmeperature with the PDO

Finally, we have Web-based tools that help illustrate the historical effects of different climate regimes on the atmosphere. The first tool that we will demonstrate is available on the monthly mean correlation web-page located at http://www.cdc.noaa.gov/Correlation/. This page computes and displays the linear correlation between an atmosphere/ocean regime and a selection of atmospheric variables. In this example, we will show the correlation between an important mode of variability called the Pacific Decadal Oscillation (PDO) and temperature. The PDO is related to long time-scale Pacific ocean variability and is associated with certain climate regimes. As can be seen, abnormally high temperatures near Seattle are associated with the positive phase of the PDO. The time-series used in the correlation page can be one of the ones supplied or one that the user supplies. One way to obtain a time-series is to use one created from the web page http://www.cdc.noaa.gov/Timeseries/. With this combination of tools, a user can obtain the time-series of precipitation over eastern Washington and then see how it relates to Sea Surface Temperatures. A similar page calculates the correlation of the U.S. climate division data with atmospheric/ocean indices.

The second example illustrates the "composite" or "average" data analysis functions. This page can be used simply to plot historical data. It can also be used to examine the average effect of some climate process (like El Niño). By doing this, we hope to remove some of the effects of year-to-year noise and make relationships show up more clearly. In addition, we can examine nonlinear effects as well as see how sensitive relationships are to the particular years used. For example, we can look at the average precipitation during the strongest El Niño years and then examine the strongest La Niña years. The results could be contrasted to those obtained from a linear correlation plot (not shown).

Using this page, we can also look at how some years compare to past years by examining rankings. Looking at February 1999, we see that the Pacific Northwest's precipitation ranked extremely high over the historical record while the southwest ranked low (a typical La Niña signal).

Our third example consists of tools and products that show in more detail effects of ENSO. ENSO is one of the largest sources of variability in the atmosphere/ocean and it has significant and large scale effects on climate. These effects are pronounced, both in the tropics and mid-latitudes, especially over parts of the United States. Part of our motivation for developing the ENSO web pages is the desire to document the historical effects of ENSO events so that current forecasts can be of use in future planning decisions. The collection of ENSO pages includes "risk" pages that show the likelihood of extreme events during different seasons and phases of ENSO over the US. For example, the adjoining figure shows the odds of having an extremely warm or cold JFM season in the US given a concurrent La Niña.

Other types of information available at our web-site are historical composites, ENSO indices, animations, reference material and a discussion of CDC's research efforts.

A summary of our interactive tools is available at:: http://www.cdc.noaa.gov/PublicData/web_tools.html
The main CDC home page can be found at:: http://www.cdc.noaa.gov/

AWRA

Symposium TOC AWRA Home page