Best Practices for Retrieving Data

Use the constraint version of the MostRecentForEachStation whenever possible

• This is considerably more efficient than the postfilter version, especially when retrieving TAF data. See the 1.2 release notes for details on the difference.

Use generated current data files

• Files (in GZipped XML and CSV format) containing the most current worldwide data for METARs, TAFs, PIREPs, and AIR/SIGMETs: http://weather.aero/dataserver/current

Or

• http://weather.aero/dataserver_current/current

Or

• http://weather.aero/dataserver1_2/current


• Please visit Current Data for more information.


• All files are gzipped and can be viewed through your browser (your browser will automatically unzip these when selected) or saved to your disk

Minimize number of requests whenever possible

• For example: when retrieving more than one station, instead of making numerous requests for single stations consider combining this into one request:

Multiple requests:
dataSource=metars&requestType=retrieve&format=xml&hoursBeforeNow=2&stationString=KDEN
dataSource=metars&requestType=retrieve&format=xml&hoursBeforeNow=2&stationString=KMIA
dataSource=metars&requestType=retrieve&format=xml&hoursBeforeNow=2&stationString=KSFO
dataSource=metars&requestType=retrieve&format=xml&hoursBeforeNow=2&stationString=KSLC
dataSource=metars&requestType=retrieve&format=xml&hoursBeforeNow=2&stationString=KLAX
dataSource=metars&requestType=retrieve&format=xml&hoursBeforeNow=2&stationString=PHNL
dataSource=metars&requestType=retrieve&format=xml&hoursBeforeNow=2&stationString=PHTO
dataSource=metars&requestType=retrieve&format=xml&hoursBeforeNow=2&stationString=CYZT
dataSource=metars&requestType=retrieve&format=xml&hoursBeforeNow=2&stationString=KORD
dataSource=metars&requestType=retrieve&format=xml&hoursBeforeNow=2&stationString=CYQB
dataSource=metars&requestType=retrieve&format=xml&hoursBeforeNow=2&stationString=CYLL
...

Single request:
dataSource=metars&requestType=retrieve&format=xml&hoursBeforeNow=2
&stationString=KDEN,KMIA,KSFO,KSLC,KLAX,PHNL,PHTO,CYZT,KORD,CYQB,CYLL, ...  

Don't unecessarily request large amounts of data

• Constraining requests as much as possible improves response time and overall server load. Consider using time, station, and geographic constraints as appropriate to make the query as specific as possible.

Data compression may be useful

• Compress data using the "compression=gzip" parameter in your request. This compresses your results and decreases the size of the returned file. XML results compress very well. This is especially relevant for users concerned with bandwidth.