1 / 8

Event Analytics: Atmospheric Science Use Case

Event Analytics: Atmospheric Science Use Case. Rahul Ramachandran Principal Research Scientist Information Technology and Systems Center, University of Alabama Huntsville Twitter : @rramachandran05 Blog: www.rramachandran.com. r ahul.ramachandran@uah.edu. Big Data Analytics: Definitions.

varana
Download Presentation

Event Analytics: Atmospheric Science Use Case

An Image/Link below is provided (as is) to download presentation Download Policy: Content on the Website is provided to you AS IS for your information and personal use and may not be sold / licensed / shared on other websites without getting consent from its author. Content is provided to you AS IS for your information and personal use only. Download presentation by click this link. While downloading, if for some reason you are not able to download a presentation, the publisher may have deleted the file from their server. During download, if you can't get a presentation, the file might be deleted by the publisher.

E N D

Presentation Transcript


  1. Event Analytics: Atmospheric Science Use Case Rahul Ramachandran Principal Research Scientist Information Technology and Systems Center, University of Alabama Huntsville Twitter : @rramachandran05 Blog: www.rramachandran.com rahul.ramachandran@uah.edu

  2. Big Data Analytics: Definitions • Are we asking the right questions? • Instead of asking: • What is the definition for “Big DataAnalytics”? • Should be asking: • Does it change the way we do science? • No – then there is nothing of value here • Yes – then what is it enabling?

  3. Traditional Analysis Process • Data – something that is directly observable therefore measurable • Knowledge – statement about a hypothesis and testing of a hypothesis is done using data • Information – measure of uncertainty about an hypothesis and the role of the data is to change the amount of information (increase/decrease entropy) Knowledge Test a Hypothesis Information Data “data by itself cannot formulate a hypothesis, rather it changes the odds in favor of or against a hypothesis” N. D. Singpurwalla, “Knowledge Management and Information Superiority (a taxonomy),” Journal of Statistical Planning and Inference, vol. 115, no. 2, pp. 361–364, 2003.

  4. Big Data Analytics: New Pathways • Data can be used to discover new anomalies that would require testing • Data can also be used extend studies limited on their use of data sets Knowledge Test a Hypothesis Formulate a Hypothesis Information Discovery (anomalies) Prediction methods Data How do we support these new pathways? • Technology • Business Model DATA-DRIVEN APPROACH TRADITIONAL APPROACH Extend studies (climatologies) Correlations Previous statement on the use of data is overly restrictive!

  5. Use Case: Events/Phenomena Studies in Satellite/Assimilation Data • Finite/bounded entity in space and time that meet some physically motivated criteria • Usually irregular • Possibly disconnected • Associated with a geolocation that may vary with time. i.e., a “track”. • Definition for event requires iterative data analysis • Examples ES events: • Heat wave • Blizzard • Derecho • Tropical cyclone • El Nino / La Nina • Tropopause fold • Urban heat island • Sea/land breezes • Blowing snow over Antartica • …

  6. Big Data Example: Event Discovery, Extending Published Studies • Event discovery query based on heuristics • Study performed using 1 year of satellite data • Can we extend such studies to 40 years of satellite data and allow the studies to be done in minutes/hours?

  7. Scoping the Problem Five common research approaches in Atmospheric Science • Event Analysis: Documenting in detail the processes for an “event” using multiple data and models • Event Climatologies: Finding representativeness of an event by capturing its characteristics such as spatio/temporal distribution, intensity, cycles, durations, correlations with other events, etc. • Synoptic Climatologies: Creating a composite evolution of an event addressing what happened before, at the time of, and after the event • Forecast Methods: Creating new methods to predict an event • Modeling Simulation Studies: Suite of simulations where the model’s physics or initial conditions are altered to evaluate their relative importance Tools supporting Event Analytics on Big Data would support four of the five common research approaches

  8. Event Analytics for Big Data

More Related