|
Visualization of seasonal-diurnal
climatology of 16th Conference on Applied Climatology, 87th Annual Meeting of the American Meteorological Society, Authors: Bjarne
Hansen, Ismail Gultepe, Supplementary webpage for poster presentation Abstract
As part of the Fog Remote Sensing and Modeling (FRAM)
project (Gultepe et al. 2006), a set of graphs has been made which displays
the seasonal-diurnal climatology of fog, visibility and related weather
variables at 198 Canadian airports, based on records of hourly
observations made during the period from 1971 to 2005 (National Climate
Archive 2003). Conditional frequencies of observed variables are graphed as
fields along two axes: time of day horizontally (hour UTC) and time of year
vertically (month), with corresponding gridlines. For additional gridlines,
local sunrise and sunset times are plotted (curved lines). Frequencies are
also plotted for additional conditions such as wind direction, precipitation,
cloud, and persistence (Martin 1972). Field values refer to probabilities of
discrete weather events and to statistics of continuous weather variables.
Additional weather variables described include: cloud ceiling, blowing snow,
snow, ice pellets, freezing rain, freezing drizzle, rain, drizzle,
temperature, relative humidity, wind speed, and clouds. The results revealed
interesting patterns in diurnal and seasonal variation of visibility. In the
marine environment, in Atlantic Canada, the maximum probability of fog is
about 35% and is likely mostly due to warm air advection (sea fog). Inland,
in fog, cloud ceiling, thunderstorm, blowing snow, IFR, BLO ALT, BLO LIMITS, snow or snow shower, snow, snow shower, ice pellets, freezing precipitation (drizzle or rain), freezing drizzle, freezing rain, rain or rain shower, rain, rain shower, drizzle, temperature average, maximum, minimum, dew point temperature average, maximum, minimum, Statistics
Each graphed point summarizes the statistics of one UTC hour of one Julian day, which gives a total of 8784 points plotted per graph. Each point is based on 31-day-running statistics for 35 year's worth data (1971-2005, or less in a few cases where station records are shorter). Thus, most points summarize 1085 pieces of data (31 x 35). Without applying a running statistic, we would obtain spiky graphs (e.g., March 1 might have 3/34 "hits", and the days before and after might each have none). The running statistic smooths out such spikes, and retains and reveals coherent patterns, such as maxima and minima. For discrete events, 31-day-running average probability is plotted. For continuous variables, three types of running statistic are plotted: 1) 31-day-running average value, simply an average of all days' data +/- 15 days of the date in question; 2) 31-day-running maximum value, the maximum of all days' values +/- 15 days of the date in question; and 3) 31-day-running minimum value, the minimum of all days' values +/- 15 days of the date in question. Caveats
The graphs should be regarded as suggestive and not as definitive. Please use meteorological discretion in interpreting them. The data that they are based on have undergone quality control; however, they still contain known flaws. For example: 1. Some airports stop making routine observations during "off-hours" at night, or switch to making of incomplete automated observations. Therefore, not all actual discrete weather events (e.g., fog, ceiling, thunderstorm, blowing snow, precipitation) are observed and recorded in the data. As a result, some graphs will display artificial discontinuities, or gaps. If you click on the asterisk beside a station's name, you will see a graph of the number (N) of observations (positive and negative) that were used in compiling statistics, and may thereby gain a sense of whether any such discontinuity is artificial. In most cases, these gaps are fairly obvious and will appear as rectangular blackout periods for off-hours periods. 2. Records for all the major airports and for most of the other airports (>90%) are at least 99% complete for the period 1971 to 2004. However, for some of the other airports records exist for only a fraction of this period. You can see how many records were used for statistics in the graphs of N. 3.
If no positive observations of a weather event
exist during a station's entire record for any day or time (e.g., thunder at
Alert, or blowing snow at 4. If no statistics exist for particular days and hours due to complete lack of observations (positive or negative), then the area of the graph for these times appears blank. 5. If relatively few statistics exist for particular days and hours due to a relatively low number of observations (positive or negative), then the area of the graph for this time will appear noisy (small areas of various colours intermixed). 6. During the period of record the making of observations has become increasingly automated. Therefore, the data is inhomogeneous over time. Interesting graphs
Each of the following graphs shows something interesting in the climatology of a particular airport. Each of the following graphs shows something interesting in the climatology of a particular airport. ·
In Montreal,
·
There is very little diurnal temperature
variation in the high arctic region of ·
In · It rains more in Vancouver than in Toronto, and it snows more in Toronto than in Vancouver. The difference in climatology of precipitation, rain and snow combined, between the cities is less pronounced than for either rain or snow considered separately: the probabilities during peak periods are similar; Toronto has a longer dry period during the summer than Vancouver. ·
In ·
In central northern · Bimodal average wind speed fields are quite common, with peaks during the spring and fall daytimes. · At Sable Island, gales are rare in July. · At Sable Island, there is a solar atmospheric tide, two daily peaks in average atmospheric pressure. · In Vancouver, there are four annual peaks in average atmospheric pressure. Fog climatology graphs
·
In ·
The peak probability for fog in the ·
Multimodal fog probability in Toronto, ·
In ·
The peak probability for fog in southwestern
BC is near sunrise during the months from August through October. In Tofino, on the exposed western coast of ·
In · In Resolute, Nunavut, fog is most frequent in July and August, coincident with the open-water season and the highest average temperatures and highest maximum temperatures, conditions conducive to formation of advection fog. Compared to fog patterns in southern locations, the seasonal peak is narrow and the diurnal peak is broad. If you see anything interesting in any graph that is not posted here, please contact me and we could add it to the list. Bjarne Hansen Cloud Physics and Severe
Weather Research Section Meteorological Research
Division Science and Technology
Branch Environment E-mail: bjarne.hansen@ec.gc.ca ReferencesFisk, Charles J., 2004: Two-way (Hour-Month) Time Section Plots as a Tool for Climatological Visualization and Summarization, 14th Conference on Applied Climatology, American Meteorological Society, 11-15 January 2004, Seattle, Washington. Gultepe,
Ismail, S. G. Cober, G. A. Isaac, D. Hudak, P. King, P. Taylor, M. Gordon, P.
Rodriguez, B. Hansen, and M. Jacob, 2006: The
Fog Remote Sensing and Modeling (FRAM) field project and preliminary results,
12th Conference on Cloud Physics, American Meteorological Society, 10-14 July
2006, Martin, Donald E., 1972: Climatic Presentations for Short-Range Forecasting Based on Event Occurrence and Reoccurrence Profiles, Journal of Applied Meteorology, 11, 1212-1223. National
Climate Archive, Documentation
for the Digital Archive of Canadian Climatological Data, Environment Phillips,
David, 1990: The Climates Of Data visualization toolsgnuplot, IDL, ploticus, Generic Mapping Tools Webpage last edited |
