Data

From Seamonster
Jump to: navigation, search


Water and data flowing: Lemon Creek discharge and air temperature, summer 2007. Correlation between warmer days and more water is due to melting on the tributary Lemon Glacier.


Lemon Creek, fed directly by glacial meltwater runoff, has increased discharge (black) after warmer weather (magenta).


Introduction

Data is good. Ancillary or metadata is better. Information is better still (ad hoc definition of Information: Useable data in context and concert with other useable data). This page describes how to get and use SEAMONSTER data, metadata, information...



DataPortal.png

  • Access (as a human) using an online user interface or web portal.
    • SEAMONSTER web portal is http://seamonster.jun.alaska.edu/browser
      • The portal format is informed directly from the database
        • ...so that new sensors show up automatically
      • The plots are for inspection purposes only, not analysis!
        • The plotting interface is zoomable and scrollable
        • Two datasets of the same type of information can be overlain on one plot
  • Metadata is at http://seamonster.jun.alaska.edu/smdisplay/info_webpage.cgi
    • Web cam portal is here.
  • Access (as a computer) using a web service.

Policy

The premise of SEAMONSTER is that all data is openly available. We're in the process of hashing out how to serve out good data safely and reliably.


  • Data can be suspect; it may not have been properly validated, cleaned up, etcetera.
    • It would be Not Good to have bad data used for some legitimate purpose or decision.
  • Server vulnerability to attack is an issue
    • We do not want to permit someone to trash the place.


food for thought: IPYDIS (http://ipydis.org) has some good points, as does the eGY, along with LTER data access policies, CEDAR "rules of the road" (at http://cedarweb.hao.ucar.edu/catalog/Rules.html ), etc..

citation: Matt Heavner


Notes

Notes sections on primary pages are deprecated...


Sources

Summer 2007 we installed several met stations, related instruments, data loggers, and some wireless. Since the focus was on instrumentation, most data recovery has been via SneakerNet, i.e. physically visiting the instruments and downloading data to laptops etc, then transporting it back to the lab (NSRL) and pushing the data into the database.


Over winter 2007-2008 we are implementing some realtime data sources that push data into the database automatically. We will use the terms Static and Dynamic to differentiate archived and real-time datasets respectively. In all cases it is important to visualize a computer database as the intermediary between the Internet and the Sensor Web.


Terminology

Time stamps are times associated with measurements. In many cases we have clusters of instruments (say temp and wind speed) making simultaneous measurements, and we often want to read these measurements as text. So CSV stands for 'comma separated values'; a common text-based way of representing multiple types of time-stamped data. The commas are generically called delimeters, and spaces or tabs can be used for this purpose as well.


Lines of text data are ok as far as they go, but data can also be made "more palatable" by formatting it in a more graphical or intuitive manner. HTML is the markup language commonly used for this purpose.


If data is desired in a compact format it is often transferred from one computer to another as binary files. Example binary file formats are jpegs (for images) or NetCDF files (for meteorological data).


CSV Example

Example from a Met-Station on the UAS Campus, April 26 2007. This is a "Comma-Separated-Value" or CSV file.


Time,TemperatureF,DewpointF,PressureIn,WindDirection,WindDirectionDegrees,WindSpeedMPH, ...
 ... WindSpeedGustMPH,Humidity,HourlyPrecipIn,Conditions,Clouds,dailyrainin,SoftwareType
2007-04-26 00:00:00,37.0,34.9,29.63,East,90,3,3,92,0.00,CLR,CLR,,WeatherDisplay:10.36, 
2007-04-26 00:06:00,37.0,34.6,29.63,North,360,0,0,91,0.00,CLR,CLR,,WeatherDisplay:10.36, 
2007-04-26 00:11:00,37.0,34.6,29.63,SW,225,0,0,91,0.00,CLR,CLR,,WeatherDisplay:10.36, 


HTML

Example from a Met-Station at the Natural Sciences Research Laboratory near the UAS campus, April 30, 2007.


Virtual globe visualizations

</big>



The Data Generation page describes where the data comes from, how it is aggregated, where the kml/etc is generated.


The OSX Database page describes the installation (and care and feeding) of the postgresql database. This details how to retrieve data and how to add a new sensor.

UAS Campus Met Station Weather Flash Notes

Here's the code to include the campus flash--I don't know if I can force the background image to campus--that'd be sweet. Right now it is some random scenic image dished out by the weather underground (wunderground).

<object width="600" height="400"><param name="movie"
value="http://www.wunderground.com/swf/Rapid_Fire.swf?units=both&station=KAKJUNEA3"
/><embed src="http://www.wunderground.com/swf/Rapid_Fire.swf?units=both&station=KAKJUNEA3"
type="application/x-shockwave-flash" width="600" height="400"
/></object>

KML

http://137.229.208.16/~heavner/SM_Beta.kml