Key points regarding this data set:
- 1961-1990 Normals (Historical average over the 30 year period 1961-1990)
- Temperature and Precipitation
- Canadian Climate Archive Data interpolated to 50 km grid
Disclaimer:
- This data is hosted on the CICS website as a public service only.
- Please refer to the cautions, limitations, and methodology sections below before proceeding.
- This data is available free for public use but remains copyright Environment Canada.
Documentation:
Data Source: For the period 1961 to 1990, all monthly total precipitation
and
mean temperature data that are stored in the Canadian Climate Archive were
interpolated to a 50 km grid, true at 60 degrees N on a polar stereographic
secant projection aligned with 111 degrees W. The regular grid of 85 rows
by
112 columns extends beyond the political boundaries of Canada. .
No data for
the neighbouring states in the United States were utilized in this interpolation
but data density along the Canada/USA border was good for the 1961 to 1990
Normals period.
Methodology: The interpolation consisted of an inverse square distance
weighting scheme applied to the data for each month. No attempt was made
to estimate missing data at climate stations. Also no adjustments have
been
made to the data to account for instrumental or other known biases. Two
passes were made through the data. The initial pass used all data within
70.8
km of a grid point to calculate the weighted (representative) value at
the grid
point. This was sufficient over much of the settled area of southern Canada
and most of Alberta. The intent here was to retain in the gridded estimate
as
much of the spatial variability in the station data as possible. The weighted
average on a 50 km grid imposes some inherent smoothing with the resulting
values being representative of 2500 km2 centered on a grid point. A second
pass was made through the grid points using a radius of influence of 500
km
for those grid points which did not have at least one station within 70.8
km on
the first pass. On the second pass, an additional requirement of a station
value in at least three quadrants was imposed. The estimates so calculated
are more smoothly varying in space and do not accurately represent the
true
spatial variability even on the 50 km grid spacing.
Limitations: Validation of the resulting gridded estimates confirmed
that the
gridded estimates of either total precipitation or mean temperature on
a
monthly basis were representative of the station data. However there is
no
elevation adjustment so the gridded values in elevated terrain such as
the
Rockies or over major uplands over the rest of the Canada are
unlikely to reflect the full orographic influence on these parameters because
few if any stations are located at or near the top of such features. If
the user’s
application is particularly sensitive to precipitation and/or temperature
over
elevated terrain, these gridded values should not be used. Canada Gridded
temperature or precipitation values should not be used in mountainous areas
unless the application is primarily to valley bottoms where the station
data
are located.
Caution: This gridded representation will be different from CANGRID
based on
the homogeneous temperature and rehabilitated precipitation data sets
developed by the Climate Research Branch of Environment Canada. These
latter data sets were developed for long-term stations by adjusting for
instrumental bias and changes in observing protocol and station siting.
There
were far fewer stations available to CANGRID than were used in the Canada
Gridded Normals but for CANGRID the more sophisticated Gandin Optimal
Interpolation technique was employed. The CANGRID values may be more
representative in mountainous regions because CANGRID is based on an
anomaly analysis from a Normals field that was estimated using elevation
as
an independent variable. As a result, a direct comparison between the Canada
Gridded data and CANGRID will result in slight differences in temperature
in
all but mountainous areas. Significant differences in precipitation will
be
apparent because of the elevation factor in mountainous regions. Significant
differences in precipitation exist everywhere in winter months because
of the
difference in the computation of snow water equivalent which affects the
determination of total precipitation. Very large differences in total
precipitation
are apparent in the north where the adjustments to the rainfall and snowfall
have had the largest impact.
Units: All temperatures are to the nearest tenth of a degree Celsius.
All
precipitation values are expressed to the nearest tenth of a millimetre.
The
decimal is explicit (e.g. 20.1 C or 46.5 mm)
For fuller documentation, please refer to the documentation provided with
the
web page.
Questions about the data or feedback should be directed to:
Ron Hopkinson
(formerly Hydrometeorologist, Environment Canada)
Custom Climate Services
3519 Queen Street
Regina, Saskatchewan
S4S 2G1
(306) 586-5489 - telephone
(306) 586-5489 - fax (phone first)
r.hopkinson@accesscomm.ca
Date Published: 2001-03-21
| 