The uppermost section of the web page contains the date and time when the information on the page was last updated (the page is set to refresh every 5 minutes), as well as the prognosis for launch (favorable, marginal, unfavorable) and a link to the latest forecast discussion written by myself or Brian. There are also links to jump to other sections of the web page.
This section contains the decoded observations from the APRS weather station nearest to the launch site and the surrounding NWS sites. Observations from these and other APRS weather sites can be found at http://www.wxqa.com/states/CO.html. Dew point temperature values are computed for the APRS sites (which report temperature and relative humidity). More information on the NWS observations can be found in the Federal Meteorological Handbook #1 (or email me if you have a question). Sun and moon information is also provided.
This section contains the winds decoded from the NWS rawinsonde observations (see http://raob.fsl.noaa.gov), and from the MAPS pseudo-sounding for the gridpoint closest to the launch site, at various levels in the atmosphere. The former are updated twice daily; the latter hourly, with about a 2-3 hour delay. The maximum wind level for each site is shaded blue
Links to various surface and upper-air charts and gridded wind fields generated in-house for the mandatory levels through 30mb. Radar and satellite links are also given.
LAPS Forecastvs Observation (if available)
This section contains side-by-side comparison of the 10km LAPS/MM5 model prediction of temperature, dew point temperature, relative humidity and winds vs. that from the APRS weather observations for the past 36 hours. The weather stations provide observations at 5 or 10 min increments, and the values used for comparison are 1-hour averages of the observations, centered on the top of the hour. The values listed for hour 12, for example, are mean values of the observations taken from 1130-1230. It is important to note that the LAPS predictions are for a level 10-20m above that of the observation site. The observation values are color coded as follows: green (0-2 kt), orange (3-5 kt) and red (> 5 kt).
LAPS Forecast vs APRS site / MAPS Analysis Statistics (if available)
In this section, the differences between the LAPS predictions and APRS observations, and between the LAPS predictions and the MAPS analyses, are statistically analyzed for the analysis (0-h) and 3-, 6-, 9-, 12-, 15-, 18-, 21- and 24-h forecasts. For each forecast time, the following statistics are provided: number of comparisons (n), mean error, RMS error (the mean magnitude of the error), standard deviation, minimum, maximum and median error. At this point, all available data are included, but addition of a user-selectable period is forthcoming.
APRS site Hourly Observation Statistics
The final section displays descriptive statistics, and an interpretive color coding, for the observations collected over the 7 days preceeding the current date and time. Observations are taken at either 5 or 10 min intervals (12 or 6 observations per hour). For various technical reasons, not all of these observations make it to the Internet. Logically, the fewer the number of observations, the less meaningful the statistics produced from these observations. In my opinion, a minimum of 2/3 of the maximum possible observations is needed for a meaningful analysis. Of most interest to us is the wind data, particularly the speed. Our objective criteria for launch is established at no more than 5 kt (6 mph), and the color coding scheme is derived from this.
Dir Persist quantifies the persistence of the wind direction for each hour. That is, how much variability is there in the mean direction? This quantity varies from 0 to 1, where 1 means the wind always blows from this direction, and 0 means the wind is equally likely from all directions (or blows half the time from one direction and half the time from the opposite direction).
Mean wind speed is, of course, important, but I give more weight to the median wind speed. The mean speed can be skewed by a few high wind speed observations, whereas the median tells me more about the typical wind speed. The same color scheme is used regardless of the number of observations for each hour: green (0-2 kt), orange (3-5 kt) and red (> 5 kt), but in the case where the minimum criteria are met (56 or more observations), the table cell is shaded the appropriate color. For < 56 observations, the text color reflects the appropriate color. The right-most column is simply a repeat of the first column (the hour), but is also color-coded. The values are shaded light gray (when the number of observations is less than 56, and bold black when 56 or more observations are included. Ideally, we would like to see both the median and mean wind speeds be under 6 kt, but a median value less than 6 gives me confidence that the speeds are typically within our contraints during the hour. Color-coding of mean wind speeds and wind direction persistence (how much variability in the wind direction over a given hour) may be added later.
Questions and comments to randall.s.collander@noaa.gov