Geoscience Reference
In-Depth Information
can be read directly. A standard rain gauge will record only
the total rain which has fallen between readings.
In many cases it is important to know when the rain
fell and at what intensity. For this purpose recording
rain gauges are used. Recent systems use a tipping bucket
of known capacity which electronically records the
number of times the bucket tips. This information can be
stored by a data logger and downloaded directly on to a
computer. Some can even be interrogated via telephone
or satellite. This is particularly useful in remote areas
where heavy rainstorms may lead to flooding lower in the
river basin.
Rain gauges are not the only means of measuring
rainfall. Weather radar systems have been developed which
can provide quantitative estimates of the rates of rainfall.
The method is based on the amount of reflection of the
radar signal from falling precipitation. Although there are
problems of interpretation of the reflected signal because
of scattering from local buildings or hills and when
snowflakes melt, it is now possible to produce maps of
areas of precipitation and their intensity, as shown in
Figure 7.11
.
In parts of the world, especially over the oceans
where there are few or no rain-gauge measurements,
satellites have been used to estimate precipitation totals
on the basis of the frequency of occurrence of the types
of cloud expected to produce rainfall (see box below).
Even the levels of outgoing long-wave radiation and
microwaves have been used in satellite estimation of
surface precipitation. Although there are many problems
it is possible to obtain an estimate of probable precipita-
tion totals in previously ungauged parts of the world.
In some countries the water equivalent of snowfall is
found by melting the snow which has accumulated in the
gauge. Clearly this is not very accurate, especially during
heavy snowfall, when a low gauge may be totally covered.
In the United States the tall gauge prevents this happening,
but the gauge tends to underestimate the amount of snow
reaching the ground. In Canada and Russia separate
snow gauges are used. There have been experiments in
measuring snow depth photogrammetrically, with aerial
or satellite photography. Where the snowfall is substantial
the depth can be obtained fairly accurately, but without
ground observations the water equivalent of the snow is
unknown.
Whichever approach is used, measurements of the
water equivalent of snowfall always entail problems and
Measuring precipitation from space
NEW DEVELOPMENTS
There are many parts of the world where it is extremely difficult to measure precipitation. Uninhabited areas presented
problems before automatic weather stations became available, but ice-covered areas, mountains and, most important,
the oceans were almost impossible to monitor adequately. Even in densely gauged areas, rainfall may vary over short
distances and not be measured properly. Now that satellites provide continuous surveillance of the globe, techniques
have been developed to allow us to estimate precipitation over most parts of the world. Admittedly the results may
not compare precisely with existing surface instruments, but where these are not available the satellite estimates
are invaluable.
Methods of estimation can be classified into two approaches: direct and indirect methods. The direct methods utilize
microwave techniques that have been available since the 1970s but have become available for precipitation estimation
only recently as the physics of the interactions became better understood. Microwave radiation can be absorbed
and scattered by precipitation particles in the atmosphere. The rate of rainfall can then be estimated from the degree
of absorption by the droplets, using a variety of assumptions. NASA plans to have a system of satellites using
microwave sensors linked to a control system with weather radar to provide calibration.
Figure 5.5
demonstrates
estimated accumulated precipitation for a particular period.
Indirect methods of estimation have a longer history, based on the types of clouds imaged by both satellite types.
Using infra-red and visible wavebands (during daylight), cloud types are identified. The probable precipitation rate and
duration can then be estimated from cloud characteristics such as the thickness, areal extent and cloud top
temperature for any given cloud type. Some cloud types such as cirrus, on their own, are unlikely to give any
precipitation at all. Cumulonimbus clouds are almost certain to produce precipitation, the amount being influenced
by the life span of the cloud, its depth, and surface temperatures.
