For a long time weather monitoring was largely a pastime of enthusiastic amateurs, but over the last century it has evolved into a well organised and professional global activity that reflects its crucial importance for a wide range of economic, environmental, civil protection and farming activities.
Weather variables such as wind speed and direction, air temperature, humidity and rainfall may all be important factors in determining the course of a wide range of events. For example, agriculture has always been heavily dependent on the weather and weather forecasts, both for its control on the quality and quantity of a harvest and its effect on the farmer’s ability to work the land or to graze his stock.
Water resources generally depend critically not just upon rainfall, but also other weather phenomenon that together drive plant growth, photosynthesis and evaporation.
Just as pollen and seed dispersal in the atmosphere are driven almost entirely by the weather, so too is the direction and distance of travel of atmospheric pollution.
Weather monitoring is also important not just in defining present climate, but also for detecting changes in climate and providing the data to input into models which enable us to predict future changes in our environment.
Because of the wide variety of uses for the information, there are a large number of environmental variables which are of interest to different groups of people. These include solar radiation, wind speed, wind direction, barometric pressure, air temperature, humidity and net radiation.
The demand for these data, usually on an hourly or more frequent timescale, has increasingly been met by the development and widespread deployment of automatic weather stations (AWSs) over the past 30 years or so.
Rainfall is mostly simply measured by accumulation in an open container, but in practice this is not necessarily so straightforward because the standard deployment of a gauge with its orifice above ground level has been found in many studies to lead to it catching too little rain.
This is due to disturbance of the airflow resulting in it accelerating as it passes over the gauge, sweeping the raindrops across the orifice, rather than falling into it.
Various designs of wind shield are available, but the best solution is generally acknowledged to be to mount the raingauge in a shallow (300 mm deep) pit about 1.5 m square such that its orifice is level with the ground surface. This prevents splash of rain into the raingauge from the surroundings. To ensure that the airflow over the ground is not disturbed, the pit is covered with a grid.
The most common way to automatically record rainfall is the tipping bucket raingauge. This has a double bucket with the divider over a pivot so that when one bucket fills and moves down the other goes up.
Rain falling on the funnel is directed through a siphon control unit and discharges as a steady stream. As each compartment fills, the bucket tilts alternately about its axis. Each tilt produces a momentary contact closure as a result of a magnet sweeping past a reed switch assembly.
A difficulty for all types of raingauge is that reliable measurements of snow are extremely difficult, owing to its propensity to drift. Heaters, either electrical or gas, can be fitted, but the former are normally too power hungry to use unless mains power is available, and the latter can be unreliable.
Air temperature is usually measured either with a thermistor or platinum resistance thermometer, the latter being more expensive, but more accurate and stable.
Air humidity is normally measured in one of two ways. The most common method relies on a variation of the wet and dry bulb psychrometer used in the traditional Stephenson Screen.
When evaporation occurs from a wet surface into an air stream the amount of cooling depends upon the relative humidity of the air and the air temperature.
Two thermometers are used, one measuring air temperature and the other kept moist by a wick trailing into a reservoir of water.
Thermometer accuracy to about 0.1 °C is required to determine the ‘temperature depression’ between the two thermometers. The method works best at air speeds more than a couple of metres per second.
At lower wind speeds, better measurements are obtained if the psychrometer is aspirated with a small fan.
Although the psychrometer is capable of giving superior humidity estimates, especially when aspirated and at high humidity levels, scrupulous maintenance is essential to keep the water reservoir replenished with distilled or deionised water, and the method does not work when temperatures fall to 0 °C.
As a consequence, an alternative ‘maintenance-free’ approach to humidity measurement is becoming increasingly popular. This uses a capacitive sensor containing a hygroscopic plastic, which changes its resistance or capacitance as it absorbs moisture from the air.
These sensors are normally housed in a louvered plastic or aluminium radiation shield and protected from the rain. They are less accurate close to saturation and they may suffer from hysteresis. They may also be prone to calibration ‘drift’ over time, especially when the air contains pollutants or high relative humidity for prolonged periods, and their lifetime is generally restricted to two to three years.
Wind measurement is usually accomplished by cup anemometers or a propeller. Rotations of the anemometer spindle are usually counted by means of a magnetically-actuated reed switch or by a digital shaft encoder.
Direction sensing by a wind vane is usually either by a potentiometer connected to the spindle or by a series of reed switches actuated by a magnet attached to the spindle.
Radiation’s most common measurement is of solar radiation. The sensor is a pyranometer which measures the power (or flux) of the incoming short-wave solar radiation in warming a flat surface, which is covered by a glass dome hemisphere to prevent cooling by wind and air currents.
The incoming radiation comprises the direct solar beam and diffuse solar radiation scattered by the atmosphere and clouds. Sometimes, these two components are required separately, which can be accomplished by a shading ring fitted to one of a pair of solarimeters.
High quality commercial radiation sensors rely on intercepting the radiant energy on a blackened flat plate containing high quality, blackened thermopile sensors, which measure the resulting temperature rise to generate a small output voltage proportional to the solar short-wave flux. Cheaper photodiode technology does not provide a spectrally flat and non-selective response across the full solar spectrum.
Energy and water balance studies need net radiation, which is the difference in total radiant energy (both short and long-wave) received from above and that reflected or emitted from the ground surface below, and measured by two identical sensor surfaces, one facing up and the other down.
As with solar radiation (short-wave) sensors the surfaces are usually protected by a transparent dome, but glass does not transmit longer wave radiation well, and so net radiometers must use relatively thin transparent polyethylene domes.
These domes are less robust and gradually degrade in sunlight, requiring approximately annual replacement, and they are also vulnerable to being pecked by birds.
A recent development in net radiometers has no domes, but two blackened PTFE-coated and slightly conical surfaces, to shed raindrops. The lack of domes reduces maintenance requirements, although the sensors are less accurate, being more prone to the variable effects of cooling by the wind, and by rain and dew on the blackened surfaces.
Photosynthetically active radiation (PAR) is that part of the electromagnetic radiation spectrum which is necessary for photosynthesis, and may be measured using photodiodes.
Constraints on instrumentation
The quality of sensors, and hence the data from them, is constrained by a number of factors, of which the most obvious is finance. In almost all cases, the sensors, data loggers or telecommunications links are required to work over a wide range of environmental conditions.
The principal problems are caused by ingress of rain, condensation and by the large temperature range (~ 60 °C) over which the equipment is expected to work.
Power requirements were once a major constraint as field instruments are generally sited far from mains power, but advances in low power sensors, solar power and battery technology have largely solved the problem for most sensors.
The power requirements of an AWS can often be fulfilled indefinitely under UK climate conditions by a 5 W solar panel and a 7 Ah battery. Many stations, indeed, operate for several months on a few dry cells. However, these power requirements demand that the power taken by individual sensors and by the logger and/or telemetry equipment is such that the average power drain is not more than about 100 mW.
Such low power requires a specialised environmental data logger. These are available at prices in the region of €2,000 for a complete system and come with relatively sophisticated software for download and data display, usually onto a laptop PC, or be interfaced to a telemetry system.
An AWS is only as good as the performance of its sensors. There needs to be a programme of regular maintenance to ensure they are operating correctly as well as less frequent (e.g. annual) calibrations.
A typical programme is:
On each monthly visit:
• Visually check sensors and leads for damage, guy ropes are secure and solar panel is clean • Radiometer domes should be clean and always visually checked for damage and replaced if necessary. If there is condensation inside then the silica gel should be replaced • The raingauge funnel should be checked for debris • Especially in summer check and top up the wet bulb reservoir with distilled or deionised water (to help preserve the conductivity of the wick) and check condition of wick Half-yearly servicing: • Inspect logger for signs of water ingress • Inspect solar panel, ensure correct orientation. Replace battery if in poor condition • Check mast is vertical with correct North orientation. Inspect condition of securing points and guy cables. Lubricate screws and fastenings • Check radiometers are mounted horizontally • Check anemometer for freedom of movement and vertical alignment • Check wind vane for freedom of movement, and vertical and northerly alignment • Check condition of water reservoir for damage (e.g. due to freezing of contents) and replace if necessary. Replace wick if necessary
• Check thermometers at two temperatures with water in thermos flasks, one containing stirred crushed ice and water (for 0 °C) and the other heated water of known temperature (use recently calibrated portable digital thermometer) • Check solarimeters against a similar sensor that has itself been recently calibrated • Check net radiometers against a similar sensor that has itself been recently calibrated • Capacitive humidity sensors can be checked in the field against a similar new sensor or taken back to the laboratory to be tested over saturated salt solutions • Slowly add a known amount of water into the raingauge funnel and note the number of tips
We are often concerned not just with individual weather variables but also with derivative measures such as the overall drying power of the air – termed the potential evaporation. This combines the temperature and humidity of the air, its wind speed, and the available radiation energy. The most reliable estimates of potential evaporation are based upon the Penman-Monteith formula. There are numerous versions, but probably the most widely used version internationally is the FAO method1.
Developments in communications technologies will allow more remote, real-time weather monitoring and access.
There are several options available for retrieving the data from the AWS:
1. Data can be manually transferred from the logger to a laptop via a download cable. 2. Alternatively, a modem built into the weather station allows you to remotely retrieve the data. 3. Weather data can be viewed on the internet with the information displayed on a public or private website (e.g. www.mea.com.au/products/weatherevap-Network).
There has recently been an explosion in the number of weather stations in operation and the ability to transmit data from them, which has helped to facilitate access to weather information beyond a limited number of stations run by the operator, to wider networks of stations.
The availability of data from spatially distributed networks has resulted in a move away from considering weather at a single point towards the evolution of weather patterns over large areas.
Many of these are operated by interested amateur owners and usually feature relatively cheap (a few hundred Euros) stations. The variables measured are normally limited to temperature, humidity, rainfall, wind direction and windspeed.
Barometric pressure is sometimes included and, more rarely, solar and/or UV radiation. Improvements in telemetry mean these data are increasingly available online.
The UK Meteorological Office provides hourly weather observations at about 125 sites (www.metoffice.gov.uk/weather/uk/observations/). In addition, more than 200 privately-operated stations in the UK and Ireland publish data online (www.weatherstations.co.uk/aws_map.htm).
The National Education Network’s Weather Monitoring System (http://weather.lgfl.org.uk/Default.aspx) is a community of UK education establishments that provides real time weather observations including temperature, humidity, air pressure, rainfall and wind for sites in England and Wales.
The quality of data from such sites is often difficult to assess, but nevertheless, they form a potentially valuable resource for many applications in which spatial and temporal information are important.
Most sites have archive data available, so that information from the past few years can be accessed. The TuTiempo website gives historic weather data for a huge number of sites worldwide (http://www.tutiempo.net/en/Climate/).
The Weather Underground website (www.wunderground.com) gives weather forecasts for locations across the world, based on national meteorological service information, and also has links to privately operated weather stations.
One of the most significant developments which has enabled automatic weather stations to become a common item in environmental studies has been the development of economical, low-power solid state data loggers.
These will continue to become more user-friendly, powerful and able to record more variables, and greater volumes of data at similar or lower prices than today. Sensors have undergone incremental improvement over the years, leading to a steady improvement in performance and reliability accompanied by a fall in price. The number and variety of sensors on weather stations is also likely to increase. Solid state anemometers, which can measure wind speed and direction simultaneously by ultrasonic time-of-flight methods, are becoming cheaper and more comparable with the costs of traditional mechanical sensors. With no moving parts, these are likely to be more reliable and maintenance-free.
The new generation of humidity sensor should be less sensitive to chemical interference, and have improved long term stability and negligible hysteresis.
This trend should continue as alternatives to other sensors are likely to appear. Visibility monitors are increasingly common on road monitoring stations. The most likely additions are those monitoring chemical variables, not only in the atmosphere (e.g. ozone, SO2 and other industrial pollutants), but also in rainfall and even in the soil.
Monitoring of current near-surface atmospheric conditions with a weather station is important, but tells us little about how those conditions arose. An important control on those conditions is the flux of water and energy, and a wide range of substances in the atmosphere. These include:
1. Water vapour
An important use of automatic weather stations in hydrology has been to determine potential evaporation – that is, the amount of evaporation which would occur from a standard crop (usually, but not always, short grass) if there is no restriction in water supply.
Very often such conditions do not exist, and the flux of water vapour from the land surface into the lower atmosphere (actual evaporation) may be very different from the potential rate. The measurement of evaporation is of vital importance to hydrologists, plant physiologists and agronomists.
2. Carbon dioxide (CO2) and methane (CH4)
These are the two most important greenhouse gases. Quantifying and understanding the release and absorption of these gases is vital in ameliorating climate change.
3. Gaseous pollutants
A wide range of organic and inorganic gaseous pollutants are released into the atmosphere.
These include soot and smoke particles, dust, finely divided chemicals and biological agents, such as bacteria, viruses and pollen.
Two techniques commonly used for measurement of the exchange of substances between the ground surface and the lower atmosphere are known as the Bowen ratio technique and eddy covariance.
The Bowen ratio method relies on an assumption that the flux of an atmospheric constituent is related to that of another through its potential (temperature or concentration) gradient. Thus, knowledge of the flux of one substance allows that of another to be calculated.
In the case of water evaporation, the ratio of sensible and latent heat fluxes (Bowen ratio) may be determined from measurements of air temperature and vapour pressure at two heights, and together with net radiometer measurements (available energy) and soil heat flux (changes in stored heat) this enables both sensible (convected) heat and latent heat (water vapour flux) to be determined.
Eddy covariance has gained in popularity in recent years. The dominant mechanism for transport in the lower atmosphere is turbulent transfer. The instantaneous flux of any substance (including sensible heat and momentum) is then controlled by its concentration and the vertical air velocity.
By integrating this flux an average vertical flux is obtained over periods of several minutes. The method requires high frequency data (ideally better than 20 Hz) and powerful computer processing.
Equipment capable of this performance, including sonic anemometers (measuring vertical fluxes in wind speed) and infra-red gas analysers (detecting infrared energy absorbed by the water vapour and hence humidity) are now commercially available. ?
Allen RG, Pereira, LS, Raes, D and Smith M (1998) Crop evapotranspiration – Guidelines for computing crop water requirements. FAO Irrigation and drainage Paper 56. Water Resources, Development and Management Service. Food and Agriculture Organization of the United Nations. Rome.
Available at: http://www.fao.org/docrep/X0490E/x0490e00.htm#Contents .
Strangeways, IR (2003) Measuring the Natural Environment.
Cambridge University Press. 534pp.
Mark Robinson and Geoff Wicks
Mark Robinson graduated from Leeds University and undertook a PhD in Environmental Hydrology at Hull University. He specialises in the impacts of land cover changes on river flows, and currently leads the long-term catchment monitoring programme at CEH, Wallingford. [email protected]
Geoff Wicks has City & Guilds qualifications in electronics and has worked in public and private sector for over 40 years, mainly in the field of environmental measurements. His main interests are low power environmental monitoring and telemetry.
The Centre for Ecology and Hydrology is a part of the UK government’s Natural Environment Research Council. It is the UK’s centre of excellence for integrated research in terrestrial and freshwater ecosystems and their interaction with the atmosphere.
Published: 01st Jun 2011 in AWE International