Tuesday, 14 January 2014

Day One & First Seminar

Day one was spent doing the usual first day things, getting an ID card, filling in forms, getting use to Linux and meeting lots of friendly new people. Despite little work been done I attended my first departmental seminar (first since back in the day) which had some relevance to the ceilometer remote sensing I did in a previous life.

-----------------------------

East East Asian Pollution Observed with Multiwavelength 
Raman Lidar in South Korea
Dr. Detlef Muller - University of Hertfordshire

   An interesting talk on how to use a specially designed Raman Lidar (Light detection and radar) to observe vertical profiles of aerosols in the heavily polluted air of Asia. Dr Muller and his team made observations up to 15km above the surface of the so called Asian Brown Cloud (ABC) which forms as a result of forest fires, un-regulated industry and windbourne dust from the deserts of central Asia. In this region the atmospheric circulation coupled with nearly constant emissions and long settling time results in wind transport of pollutants across international boundaries impacting on air quality far from source.
  Unlike traditional Lidar that are used in boundary layer research that report backscatter coefficient to identify concentrations of aerosols, the Raman Lidar's data can be used to determine the extinction coefficient of the atmosphere (useful for global climate modellers) and to derive the particle depolarization ratio which can be used to infer the shape of the particle. Using all three properties Dr Muller's team were able to identify qualitatively possible source areas/regions of the observed air pollution at different heights in the atmosphere based on known properties about the size of certain aerosols. A common observation made by the team was a low level layer of local air pollution above which was a layer of 'clean' aerosol free air capped at around 5 km with windborne dust from the deserts of central Asia or industrial emissions from China. However there is a degree of uncertainty when multiple sources of aerosol may be present over the Lidar. An interesting finding was that the optical depth of the atmosphere in East Asia from observations in South Korea and China ranges from 1 to 1.5 which is incredible when compared with the values ~0.2 observed in Europe where much stricter air pollution controls are in force.
   The final section of the seminar focused on using the data to infer even more information about the types of aerosol observed (e.g. Quartz dust from deserts using SiO2 (Silicon dioxide)) by using complex mathematical analysis techniques. This early research (Dr Muller estimated that to improve techniques and reduce uncertainty to the levels required by modellers could take 10-15 years) showed that by using distinct Raman lines (the signature of some aerosol have a distinct wavelength that they scatter at) that they could also determine the elemental composition of the observed aerosol layers.
   In summary (and in response to questions) Dr Muller stated that to improve the algorithms there needed to be advances in numerical analysis and dedicated observation campaigns, to get in-situ data to verify the Raman Lidar observations and numerical solutions used to identify aerosol species, in which remote sensors and sampling teams needed to work in close unison to a set of common goals.

--------------------------

I am aiming to take away one lesson from each seminar that I attend over the coming years. Today I learnt that I need to embrace uncertainty and accept the limitations of techniques used. This is especially true as I am to learn many new techniques over the coming years and may revisit subjects of old (Mixing Layer Height determined using Ceilometers) in which my thinking was influenced by negativity towards the accuracy and uncertainty of the techniques used.   

No comments:

Post a Comment

Note: only a member of this blog may post a comment.