Papers by Sara Zandi
Posters by Sara Zandi
Conference Presentations by Sara Zandi
The impacts of atmospheric aerosols on public health, climate, and atmospheric composition are cu... more The impacts of atmospheric aerosols on public health, climate, and atmospheric composition are currently among the most widely studied topics in environmental science research. Of most interest to its inhabitants, near-ground atmosphere (up to 1km AMSL) is most likely to influence the activities of our world. Short-term forecasting of air quality can help to avoid excessive medication, reduce the needs for hospital treatments, and even prevent premature deaths by placing restrictions on traffic and industry.. Artificial Neural Networks (ANNs) provide a powerful platform for air pollution modelling. The research described in this paper is aimed at predicting maximum PM10 concentrations for a day using the meteorological parameters for that day and from the previous day. An Artificial Neural Network was used to estimate maximum PM10 concentrations in Timaru, New Zealand. Despite having a low population, Timaru is known to suffer from degraded air quality during the winter. Meteorological and PM10 data were obtained from Environment Canterbury (Ecan, 2015) air quality monitoring stations located at ANZAC Park and the industrial suburb of Washdyke (6 km apart). The input variables selected for input into the model were mean and wind direction (º) and speed (m s-1), mean temperature (ºC) at 2 and 6 meters, mean relative humidity and, the mean and maximum PM10 values for the previous day. Mean wind direction and speed, and mean temperature (ºC) at 2 and 6 meters for the day on which to estimate PM10 were also used. The input and target variables were normalized between 0.1 and 0.9 and then presented to the network.
Thesis Chapters by Sara Zandi
GeoComputation is a relatively recent and emergent discipline containing numerous methods and tec... more GeoComputation is a relatively recent and emergent discipline containing numerous methods and techniques for fields such as electrical engineering, computer science, geography, biology. This thesis set out to describe the development of the research domain for GeoComputation and learning area so, to compare and evaluate the methods conventionally used for problem definition and investigation by those working in the discipline.
Books by Sara Zandi
Uploads
Papers by Sara Zandi
Posters by Sara Zandi
Conference Presentations by Sara Zandi
Thesis Chapters by Sara Zandi
Books by Sara Zandi