Potter T.D., Colman B.R. (co-chief editors). The handbook of weather, climate, and water: dynamics, climate physical meteorology, weather systems, and measurements
Подождите немного. Документ загружается.
chemicals. The adhesive used in the substitute brand was dissolved by the lacquer
thinner, which in turn contaminated the mirror.
The hygrothermometer sensors are housed in an enclosure to protect them from
elements like precipitation, solar radiation, dirt, and insects. Any of these could
cause false readings. The shelter also includes a fan that ensures a free flow of air
over the sensor. Even with these precautions, the shelter is subject to small variations
in temperature caused by solar radiation that varies from day to day and from place
to place. Varying wind speeds also affect the amount of airflow through the shelter,
which can cause temperature changes. In addition, the electronic components can
cause a small calibration drift. Because of these factors affecting the temperature
readings, the hygrothermometers are checked weekly using liquid-in-glass
thermometers.
Even though these instruments at airports are not meant to meet climatological
requirements (there are separate networks for those), these data are used as an index
of change for the local area. In addition, airports are notoriously bad for climato-
logical measurements because of the heating effects of runways, roads, buildings,
and moving aircraft.
3 TEMPERATURE SITING STANDARDS
Temperature sensors should be located over terrain (earth or sod) that is typical of
the area around the station. Unfortunately, some thermometers are located on roof-
tops for security purposes. Rooftops are not desirable locations for measuring air
temperature. The siting problem is the main source of the bank thermometer
phenomenon, whereby surface air temperatures are frequently reported too warm.
Figure 5 Hygrothermometer interior showing intake air screen. Finned object on bottom is
Peltier device. See ftp site for color image.
726
INSTRUMENT DEVELOPMENT IN THE NATIONAL WEATHER SERVICE
Because the sensing element of a thermometer absorbs more radiation when exposed
to the sun, than the air itself, it must be shielded. The shielding also protects the
sensing element from precipitation, dirt, and insects, any of which could cause a
false reading.
The preferred temperature sensor height is about 2 m (5 ft) above the ground (eye
level). This height needs to be adjusted in areas that accumulate deep snow cover.
Sensors located too close to the ground will read too low for minimum temperat ures
and, conversely, maximum temperatures will read too high during the day. Sensors
mounted on towers may be unrepresentative because temperatures can vary greatly
in the lowest levels of the atmosphere near the ground. They can be especially
unrepresentative during clear, calm mornings with inversions, where temperatures
can vary 8
C (14.4
F) or more in the first 6 0 m (200 ft) above the g round.
Temperature sensors should be installed in a position to ensure that measurements
are representative of the free air circulating in the locality and not influenced by
artificial conditions such as large buildings, cooling towers, or expanses of concrete
and tarmac.
4 PRECIPITATION
Many centuries ago, the people of the Middle East measured precipitation with
buckets and measuring sticks. The data were used to levy taxes since precipitation
was associated with agricultural output. Although the number of measuring tech-
niques has increased, the simple bucket and measuring stick prevails in numbers over
any other methodology used throughout the world.
The NWS uses the 20.3-cm (8-in.) (size of the orifice) precipitation gauge for
manual observations. See Figure 6. The preci pitation falling through the orifice is
Figure 6 Manual precipitation gauge, 20.3-cm diameter. See ftp site for color image.
4 PRECIPITATION 727
funneled into an inner tube whose cross-sectional area is one tenth that of the outer
can. This ratio provides the magnification of the measuring resolution by a factor of
10. If the precipitation exceeds the total capacity of the measuring tube (2 in.), it
spills over into the overflow can and must be poured into the receiving tube to be
measured. For solid precipitation, the collector funnel and receiving tube are
removed and the precipitation is caught in the overflow can. The catch is then
melted to determine the water equivalent of the precipitation. This gauge also
serves as the standard to which other NWS gauges are compared.
Unfortunately, such simplicity could not serve all of the NWS’s requirements. In
the latter part of the nineteenth century, the tipping bucket gauge was developed
to record rainfall rates and amounts. See Figure 7. Rainfall was funneled from a
12-in.-diameter orifice through a small spout to a tipping bucket mechanism. This
mechanism consisted of two buckets. The par ticular bucket under the spout filled to
capacity, lost its balance, and tipped. As it tipped, it activated a mechanical counter
to record the total rainfall. See Figure 8. In later years, the tip caused the closure of a
mercury switch, which sent an electronic impulse to an electronic counter. With the
concern over the potential hazard of mercury, the mercury switch is being replaced
with a reed switch. The number of tips recorded over a period of time determined the
rainfall rate. The rainfall amoun t was recorded in a collector tube beneath the gauge.
The need to separate the means of measurement of rate and total amount is caused by
splashing at the tipping mecha nism during higher rainfall rates.
Another limitation of the tipping bucket gauge is that it does not work accurately
with solid precipitation. A heated version was tried, but the heat caused further
inaccuracies by causing evaporation and, in some cases, a thermal plume that
prevented all the snowfall from falling into the gauge.
To catch all types of precipitation, the first of a number of weighing gauges was
developed. See Figures 9 and 10. This ‘‘universal’’ gauge caught precipitation falling
Figure 7 Tipping bucket rain gauge. See ftp site for color image.
728
INSTRUMENT DEVELOPMENT IN THE NATIONAL WEATHER SERVICE
through a 20.3-cm (8-i n.) orifice and was d irected into a galvanized bucket on the
platform of a spring-scale weighing mechanism. The weight of the precipitation
depressed a scale and through a mechanical linkage deflected a pen arm across a
clock-driven rotating-paper chart. The most commonly used version of this gauge
has a capacity of 12 in. of liquid precipitation and a 24-h recording period. For
increased resolution, not accuracy, the recording is expanded across a dual traverse
of the pen arm. During the winter season when snow, ice, and freezing temperatures
are likely, the gauge is winterized with an anti freezing solution to prevent damag e to
the measuring bucket. Other precipitation gauges have been developed that measure
Figure 8 Tipping bucket mechanism. See ftp site for color image.
Figure 9 Universal precipitation gauge with wind shield. See ftp site for color image.
4 PRECIPITATION 729
the amount of precipitation collected in a container. The amount is determined by
technologies such as strain gauges, shaft encoders, and load cells.
5 PRECIPITATION STANDARDS
The gauge should be mounted so the orifice is in a horizontal plane. The height of
the orifice should be as close to the ground as practicable. In determining the height
of the orifice, consideration must be given to keepin g the orifice above accumulated=
drifting snow and minimizing the potential for splashing into the orifice. The
immediate surrounding ground can be covered with short grass or be of gravel
composition, but a hard flat surface, such as concrete, gives rise to splashing and
should be avoided.
The catch of a precipitation gauge should represent the precipitation falling at that
point. This is not always attainable because of the effect of wind, and thus care
should be exercised in the selection of a precipitation gauge site to minimize the
wind effects. Towers and rooftops are poor locations for precipitation gauges.
Precipitation gauges should be located on level ground at a distance from any
object of a minimum of two, and preferably four, times the height of the object above
the top of the gauge. An object is considered an obstruction if the included lateral
angle from the sensor to the ends of the object is 10
or more. Beyond this range,
objects that individually, or in groups, reduce the prevailing wind speed, the tur-
bulence, and eddy currents in the vicinity of the gauge may provide a more accurate
catch. Thus, the best exposures are often found in orchards, openings in groves of
trees, bushes and shrubbery, or where fences and other objects together form
effective windbreaks. Rain gauges should never be installed on roofs or towers
because of increasing winds with height and the presence of eddy currents.
Figure 10 Weighing bucket in universal gauge. See ftp site for color image.
730
INSTRUMENT DEVELOPMENT IN THE NATIONAL WEATHER SERVICE
In order to reduce losses caused by wind, a windshield is recommended to be
installed on gauges where 20% or more of the annual average precipitation (water
equivalent) falls as snow.
6 WIND
From earliest times, attempts have been made to measure the effect of the speed of
the wind. The wind force was referenced to the effect it had on objects. Without
realizing it, people of old used a form of the Beaufor t wind scale without the
Beaufort number or the speed of the wind. The wind direction was something
more obvious even without instruments. References to wind direction can be
found in the early books of the Bible.
Somewhere in time, an early scientist must have tried to associate the revolutions
of objects like a windmill, in a given amount of time, as a measure of wind speed. In
the seventeenth century, seamen estimated the wind by the angular movement, from
the vertical, of a flat plate that rotated on a horizontal bar. This was followe d by the
mounting of objects on rotat ing wheels to catch the wind. Some of the wind catchers
were made of sail cloth, and different shaped wooden and metal disks.
By the time the federal government got into the weather business, the rotating cup
anemometer was the most prominent method to measure wind speed. For many
decades, meteorologists experimented with various anemometers. It was then
known that the rotational speed of the cups was a function of wind speed and the
density, viscosity, and turbulence of the air. What was also important was the
diameter of the cups, cup shape, the number of cups, arm length, the moment of
inertia of the system, and the effect of precipitation. By the late 1920s the four-cup,
10-cm (4-in.) diameter, hemispherical-shaped cups were replaced by the three-cup,
12.5-cm (5-in.) diameter, hemispherical-shaped cups. A few years later, the
hemispherical-shaped cups were repla ced by semiconical cups. See Figure 11.
The semiconical cups did not overestimate gusty winds as much as the hemispheri-
cal. Throughout these transitions, the NWS made correction tables available so the
speeds of the different systems could be compared. Other than for the methods of
recording wind speed and direction, this system has remained essentially the same.
Early wind recorders were composed of worm and toothed gears that indi cated
the passage of a mile of wind. Subsequent to this, electrical contacts were used to
measure miles of wind. These miles of wind were indicated by buzzers, blinking
lights, or marks on a chart. Wind speed could then be determined by the amount of
time it took between consecutive contacts. To improve the resolution of wind speed
measurement, contacts were eventually made for every 1 =60th of a mile.
This system was eventually replaced by a direct reading anemometer. See
Figure 12. It contained a magneto or small electric generator using a permanent
magnet. A spindle is connected to the armature of the magneto. The revolutions per
minute of the spindle, caused by the rotating cups, determine the amount of electrical
current generated by the magneto. See Figure 13.
6 WIND 731
In the 1980s, the internal electronics were changed. Light from a light-emitting
diode was pulsed by slits in a disk that rotated around the spindle in the anemometer.
This ‘‘light chopper’’ replaced the magneto and provided improvements. One such
improvement was reducing the starting torque with the removal of the magneto.
Another improvement was the elimination of questions regarding time constants
with the generator system. Time constants with the former system had to be esti-
mated because the dial indicator and gust recorder have time const ants determined
by inertia. The output of the light chopper is simply the numbe r of light counts per
second.
Figure 11 Rotating cup anemometer and wind direction sensor atop 10-m tower. See ftp site
for color image.
Figure 12 Wind speed and wind direction indicators. See ftp site for color image.
732
INSTRUMENT DEVELOPMENT IN THE NATIONAL WEATHER SERVICE
Wind sensors should be oriented to true north. The site should be relatively level,
but small gradual slopes are acceptable. The sensors should be mounted on a freely
exposed tower, 10 m (30 to 33 ft) above the average ground height within a radius of
150 m (500 ft).
7 PRESSURE
In the seventeenth century, Torricelli balanced the pressure exerted by the atmos-
phere against the weight of a column of mercury with a vacuum above the mercury
column. The height of this column was approximately 760 mm (30 in.). This is
basically the same technology still being used today. These terms of height are
still being substituted for pressure units in certain applications, such as aviation.
Synoptic pressure analyses use the hectopascal, which is a recent standard change
from the millibar.
The NWS has been phasing out the use of the Fortin tube to determine atmos-
pheric pressure. In this device, a level of mercury in a cistern is raised by turning a
thumbscrew to a zero scale. The zero scale is an ivory tip. Just before the top of the
mercury in the cistern touches the ivory point, the image of the tip can be seen in the
surface of the mercury. When the tip and its image just touch, or the mercury is
‘‘dimpled,’’ the proper level has been reached. The height of the mercury is then
measured by a vernier scale. Corrections are made for gravity and the expansion of
glass and mercur y with temperature.
With the advent of aviation, pressure readings were required more often. The
reading of the mercury column was not something that could be done quickly, so the
aneroid barometer was used. With the aneroid, the pressure of the atmosphere is
balanced against the force of springs in an evacuated chamber made of metal. Later
Figure 13 Wind gust recorder. See ftp site for color image.
7 PRESSURE 733
on, the construction of the metal bellows chamber itself provided the balancing
force. The motion of these chambers were exaggerated by levers to a dial. See
Figure 14.
Permanent graphic records of the pressure values were provided by using a pen
arm that made a trace on a clock-driven drum. This recording device was called a
barograph. Aneroids had to be calibrated against the mercury barometer periodically
and differences were posted on the aneroid as a correction factor.
With the concern about the dangers of exposure to mercury from a broken
mercurial barometer, these barometers are being replaced by pressure transducers.
In these instruments, the atmospheric pressure is still balanced against a bellows, but
the pressure is measured by the force on a resonator. The frequency of oscillation of
a crystal quartz resonator varies with the atmospheric pressure. Corrections are made
for pressure changes caused by temperature. The NWS has found that these instru-
ments are very stable and are being used to replace the former aneroids and the
mercury reference barometers.
Pressure-sensor-derived values are of critical importance to aviation safety and
operations. This is one of the parameters that cannot be sensed by humans other than
by the rapid rise or fall of pressure. This is usually experienced by ‘‘ear popping’’ or
sinus pain.
Care should be taken to ensure that pressure sensor siting is suitable and accurate.
The elevations of the sensors shall be determined to the nearest foot. Pressure
sensors are usually located indoors. Sensors should not be exposed to direct sunlight
or in drafts from heating and cooling. With pressure tight buildings, each press ure
sensor should be individually vented to the outside to avoid pressure variations due
to ‘‘pumping.’’ These pumping or pressure variations occur with the cycling of
heating and air-conditioning systems .
Figure 14 Altimeter setting indicators. See ftp site for color image.
734
INSTRUMENT DEVELOPMENT IN THE NATIONAL WEATHER SERVICE
8 CLOUD HEIGHT EQUIPMENT
The height of clouds is most important for aviation. Earliest measurement methods
were by the use of ceiling balloons. Balloons of various weights, typically 10 g with
a nozzle lift of 45 g or a 30-g balloon with a nozzle lift of 139 g, were inflated with
helium. A watch was used to time the interval between the release of the balloon and
its entry into the base of the clouds. The point of entry for layers aloft was consi-
dered as midway between the time the balloon began to fade until the time the
balloon completely disappeared. With surface-based clouds, the time interval
ended when the balloon completely disappeared. During the day, red balloons
were used with thin clouds and black balloons were used with thicker clouds. At
night, a battery-powered light was attached.
In the 1930s, the beam from a ceiling light was projected at a 45
elevation angle
into the sky. The projector was rotated about the vertical axis until the ligh t beam hit
the lowest cloud. The observer then paced off the distance from the projector to
where he was directly under the cloud hit. With the geometry of this scheme, the
paced distance equaled the height of the cloud. It soon became obvious that it would
be less time consuming to project the light vertically into the air. See Figure 15. At a
specified baseline away, a clinometer was used to measure the angle between the
ground a nd the spot of light on the cloud. Knowing the baseline length and elevation
angle in this right-triangle situation made it easy to determine the height with a
lookup table. Clinomete rs were used to determine elevation angles. This instrument
had wire cross hairs at one end and a narrowed neck at the siting end. This shape
gave it the name ‘‘beer bottle.’’
The human detector end of this scheme was later automated by the use of a
photocell that scanned the vertical path until the spot of light on the cloud was
observed. The projector light was modulated so the photocell would only pick up
Figure 15 Fixed-beam ceilometer. See ftp site for color image.
8 CLOUD HEIGHT EQUIPMENT 735