Data from a VHF wind-profiling radar are being used as a teaching aid in the Department of Meteorology at the University of Reading. An example is given of data obtained during the passage of a train of baroclinic waves which shows how the radar, together with output from weather-forecast models, can vividly illustrate processes occurring over a range of scales.

A volcanic ash detection product using AVHRR data has been developed for use in the London Volcanic Ash Advisory Centre, operated by forecasters at the Met Office. The product is an image that shows the brightness temperature difference between two infrared channels, which can be used to discriminate between clouds containing volcanic ash particles and those containing water droplets or ice crystals. Factors such as water vapour, water droplet and ice crystal content, opacity, cloud top temperature, ash particle size, surface emissivity properties and instrumental effects all have an effect on the spectral signal represented in the image. AVHRR imagery has been used to study the volcanic clouds ejected from eruptions of Mount Etna (Italy) in July 2001, Hekla (Iceland) in February 2000 and Mount Spurr (USA) in September 1992 to demonstrate the application of this product during volcanic events. The volcanic ash detection product provides essential information that can help forecasters locate volcanic ash and give appropriate guidance through advisory statements to the aviation industry and thus help avoid dangerous and expensive encounters between aircraft and volcanic ash.

With growing international collaboration (e.g. the European Composite Observing System programme – EUCOS), changing demands of data users, technical progress with new measuring systems and national budget cuts, the design of meteorological networks has become increasingly important. By analysing future user requirements and taking into account cost-benefit considerations, MeteoSwiss has elaborated a network configuration for ground-based and upper air stations. It is evident that for ground-based stations no separate networks should be distinguished – instead, the needs of climatological modellers and forecasters should all be integrated into one network. The future upper air network will be complemented with additional windprofilers, a mini radar and additional radar information supplied by neighbouring countries.

Currently the South African Weather Service (SAWS) compiles seasonal rainfall and temperature outlooks by combining output from various models in a subjective manner. A group of scientists discusses the forecasts from a variety of models and then decides, from these discussions, what the strongest climate signal is. A map is drawn to represent the ‘gut feeling’ from the specialists based on the results from the various models. Although the models used are all scientifically sound, the compilation of the outlook remains subjective. In order to compile a single probability forecast out of the many different model outputs an objective method has to be implemented. This paper proposes such a method. The results were tested during the December–January–February season over a retroactive period from 1991/92 to 1999/2000. Results show that a combination of different models consistently delivers a more skilful forecast than any individual model on its own.

Artificial Neural Networks (ANNs) that were originally proposed to model brain functions mathematically are now being used in many branches of research, including the atmospheric sciences. The main contribution of this paper is the development of ANN to identify the presence of thunderstorms (and their location) based on microwave measurements from all five AMSU-B frequencies and on surface observations over Iran (45–65°E, 25–45°N). From the surface observations, the position of each of the different meteorological events (thunderstorm, heavy rain, rain, light rain, snowfall and clear sky) is located in AMSU-B images. An area around this position with roughly the same brightness (visually) is selected. Then the brightness temperatures (TBs) of a few randomly selected pixels from this area are noted. In all, there were 560 pixel-points, over 20 passes of AMSU-B, in these selected areas. It is found that the thunderstorms are related to image features that are bright in all of the five-frequency images (i.e. have low TB), as expected.

The architecture of the neural network that was finally selected for identifying thunderstorm events (as against other reported events), after a series of trial and error for the best accuracy, was achieved using the four-layered Feed-Forward Artificial Neural Network (FFANN) with five nodes in input, two hidden layers with four nodes each, and one node in the output layer. In the ANN output, the thunderstorms clustered clearly away from other phenomena. ANN gave more correct thunderstorm recognition (96.7%), whereas the discriminant analysis gave 95.3% correct recognition without clear separation between the two classes.

Formation of frost on paved surfaces presents a potential hazard to the motoring public in cold climates. Temperatures of the paved surface are not measured routinely by the National Weather Service and are not part of public forecasts of winter conditions, yet highway maintenance personnel must make frost suppression and anti-icing decisions based on expectations of future paved-surface temperatures. The Road Weather Information System measures road surface, air and dew-point temperatures, road surface conditions, and wind data at numerous locations in the state of Iowa and reports the data in real-time to maintenance offices. A model based on simple concepts of moisture flux to the surface was developed that uses data from roadway weather stations or forecasts of dew-point temperature, air temperature, surface temperature and wind speed to calculate frost accumulation on bridge decks in Iowa. The analysis showed that the model has sufficient accuracy to be used as an operational tool for assessing frost accumulation on bridge decks. A logistical regression procedure was developed to determine the probability that a maintenance worker will observe frost for a given calculated frost depth.

A new method of constraining divergent attenuation corrections for weather radar systems is presented. This was motivated by the need for reliable attenuation corrections when making quantitative precipitation estimates using a small, mobile X-band radar at short range. The approach is suitable for systems requiring attenuation correction in real-time and requires no auxiliary data. An outline of the literature on attenuation and its correction for single-polarisation weather radar is presented. The traditional form of correction is known to be problematic due to the divergence, which may occur in its estimation. The form of constraint presented is based on the representation of attenuation correction originally derived for space-borne radar configurations. This method determines when divergent estimates will occur and allows a more realistic application of the constrained correction. Examples of attenuation correction applied to X-band radar observations in comparison with ground clutter returns are presented, showing good agreement and hence good absolute calibration. Selected profiles are then used to determine the influence of measurement errors on attenuation estimation in the context of this representation. The paper concludes with a discussion of the practical limitations and considerations in applying an attenuation correction to quantitative weather radar rainfall estimates.

Orographically enhanced rainfall is the dominant climatic feature of the alpine regions of South Island, New Zealand. Most of this rain occurs in strong north-westerly winds although a forecast difficulty exists in that near-saturated north-westerly flows do not always produce heavy rain. Air masses impinging on the Southern Alps are generally carried over the alpine divide in times that are normally regarded as insufficient for clouds to produce precipitation-sized particles. Hence a possible explanation for these observations is that precipitation occurs only when pre-existing clouds advected onshore contain cloud drops that are sufficiently mature to allow for raindrop growth in the available time. A model combining advection of moist air over the topography and IR satellite data was used to show that synoptic cloud is necessary for the formation of heavy rainfall in the Alps.

The pricing of weather derivatives motivates the need to build accurate statistical models of daily temperature variability.Current published models are shown to be inaccurate for locations that show strong seasonality in the probability distribution and autocorrelation structure of temperature anomalies. With respect to the first of these problems, we present a new transform that allows seasonally varying non-normal temperature anomaly distributions to be cast into normal distributions. With respect to the second, we present a new parametric time-series model that captures both the seasonality and the slow decay of the autocorrelation structure of observed temperature anomalies. This model is valid when the seasonality is slowly varying. We also present a simple non-parametric method that is accurate in all cases, including extreme non-normality and rapidly varying seasonality. Application of these new methods in some realistic weather derivative valuation examples shows that they can have a very large impact on the final price when compared to existing methods.

We discuss how weather forecasts can be used in the pricing of weather derivatives and derive results for the most important types of weather index and contract. We show that calculating the expected payoff of linear contracts on linear indices requires only forecasts of the mean temperature over the contract period. Calculating the expected payoff of linear contracts on non-linear indices requires forecasts of both the mean and the distribution of temperatures, but not of the dependence between temperature distributions on different days. Calculating the expected payoff of non-linear contracts requires forecasts of the full multivariate distribution of temperature over the whole contract. For contracts that extend beyond the end of available forecasts, correlations between the forecast and post-forecast periods must be taken into account when estimating this distribution. We present two methods by which this can be achieved, both of which combine information from climatological models of daily temperature with information from probabilistic forecasts.