The final update for all KARL processes!

• Dismantled antenna and ELLIOTT system, terminating data collection after a full year of data collection.

• Completed Kolorgramme and Nogata for all data collected.

• Collected data in Colorgramme form is still available on RMOB.  Go to their "archives" for the appropriate month and look for the data - K.E. Slovacek

• Added Kolorgramme for the month of July (v1.1c)

• Added Nogata entry for the month of July

• Added Kolorgramme for the month of June (v1.1c)

• Added Nogata entry for the month of June

Nogata is finally here!  The next generation of Kolorgramme and Duratogramme, Nogata is the official statistical tool of the KARL project.  However, Nogata is much more than a tool for just the KARL project, as now it is open to any radiometeor astronomer who is interested in redefining how they observe and visualize data.  Nogata aims to be a user-friendly, universal application for MBC data, and ultimately hopes to lower the barrier to entry for the field as a whole.

Features

• Core functionalities:

  • • Daily data plots for Diurnal Variation and Duration

    • Average day creation, with outlier detection

    • Polynomial Curve fitting

    • Sine Curve fitting

    • Duration Percent of Population pie chart

    • Hourly meteor forecast (sine and poly)

• Cloud-ready Google Sheets version

• Documentation and User Manual

With this in mind, Duratogramme is officially being decommissioned after the month of April, and all further KARL v2.x data will be available in Nogata.  However, due to the fact that KARL v1.x is not yet running in Spectrum Lab, it is not compatible with Nogata, and therefore, will continue to be updated in Kolorgramme until it is ported over to Spectrum Lab.

• Added Kolorgramme for the month of April (v1.1c)

• Added Duratogramme for the month of April (v2.0)

With an improved MATLAB script for parsing RMOB text data into a format that is easily importable into Kolorgramme, data will continue to be added to Kolorgramme from the v1.1 run on the first of every month.

Data collection for Kolorgramme halted indefinitely to make time to further KARL v2.x development.  RMOB will continue to log and publish data from the KARL v1.1 run, but there are no further statistics available on Kolorgramme after this date.

Studying Duration

• It's time!  KARL branches out from studying diurnal variation of all meteors, to studying duration of meteors from a given population.

• To do this, a group of Spectrum Lab utilities were added as a parallel process to KARL 1.1.

  • One of these acts as a filter to mitigate unwanted noise from nearly continuous tones and other sources.  (referred to as the Spectrum Lab Filter)

  • The other acts as the detector that counts meteors and logs their duration.  (referred to as the Spectrum Lab Meteor Detector)

• Soft transition into Standard Build, data between Beta and Standard Build are fully comparable

• Data collected via KARL 2.0 is extremely different from data collected in previous versions, and as such, new data analysis and statistical methods are required.

• Using a spreadsheet and graphing utility (Duratogramme), the number of meteors detected at a particular duration are collected and displayed in a few different ways.

  • Duratogramme-H: A prediction curve for number of possible packets transmitted at different times of day.  Version H uses the generated curve fit formula from the KARL 1.1a subset.

  • Duratogramme-DS and Duratogramme-DK: A representation of the variance from day-to-day of what percentage of the population of meteors was of each length, measured in normalized time slices (DS), and possible packets transmitted (DK).  

  • Duratogramme-EK:  Compares measured meteor population to a decaying exponential function.  Uses data by time slices.

  • Pie Chart: Illustrates the breakdown of the total population of meteors by duration measured in possible packets transmitted.

• Statistical data is not available at this time, as there is no established data to compare this work to. (wow!)

Technical Summary

• 01/17/18: From Alpha

  • Formatted data output file

  • Initialized Duratogramme post-processing

  • Updated Detection Threshold

Technical Summary

• 12/27/17: Initial commit

• 01/03/18: Added SL Filter

  • FFT Autonotch Filter

  • FFT Bandpass Filter

• 01/05/18: Updated Detection Threshold

• 01/08/18: Updated Detection Threshold

• With the future observation that the peak started to trend towards later hours (likely due to noise), this date was selected as a cutoff point for curve fitting using KARL 1.1. 

• Internally noted as the KARL 1.1a subset cutoff date.

• No parameters were changed, data is fully comparable to future data collected

• During an installation procedure to prepare for KARL 2.0, something caused HROFFT to crash.

  • Reason for failure currently unknown

• HROFFT and Colorgramme were restarted as soon as the problem was noted, resolving the issue

• The system was not operational for 1.5 hours (1400-1530); no data was collected during this time.

• With new data being collected via KARL v1.1, there's now room to add a meaningful statistical analysis.

• Using a custom Spreadsheet-Colorgramme hybrid (Kolorgramme, available in the header), each hour is being averaged across days, and the resultant "average day" is being used for most statistics.  

• Innate to Kolorgramme:

  • A ratio of the maximum to the minimum (for the average day and per each day), which can be compared to the goal ratio of .25 using percent error.  

  • Standard deviation is being computed per row (hour).  This is being used to mark data in the Kolorgramme that falls outside one (italics), or two deviations (blacked out).  Data outside two deviations is omitted from further computations (dynamically updated to reflect changes in the average with all data)

  • The standard deviations are also being combined via Root Mean Square, which produces an "Average Variance".  This can be used as a measure of data precision.

• Analysis in MATLAB is planned for the future, including a general curve fit, and comparisons to other real data from similar radio meteor stations.

• At a routine check, the system was found to have been rebooted and returned to the logon screen sometime in the previous night.  

  • Reason for failure: At 0302, a vital Windows security update was pushed to all systems, bypassing settings to "queue" updates, and restarting Windows 7 computers - including Elliott.

• After logging on, KARL processes were restarted with no errors at 0650.

• The system was not operational for ~4 hours (0310 - 0650); no data was collected during this time.

• At a routine check, it was noted that Colorgramme has "drifted" out of sync with the generation of new HROFFT images. 

  • Reason for failure currently unknown.

• In an attempted fix, both programs were restarted.  During the restart, an error occurred, causing Colorgramme to "skip" an hour, and look for a file that did not yet exist.  

  • Reason for failure currently unknown.

• All programs were terminated, and the system was re-synced with Dimension4, which resolved the issue

• The system was not operational for 20 minutes (1550 - 1610); no data was collected during this time.

• Machined components to match Moxon Table outputs

• Assembled antenna according to CAD and Moxon Table

• Developed and machined mounting solution (PVC boom)

• Mounted antenna to the banister

• Starting from an established pattern proposed by LB Cebik's paper "A Simple Fixed Antenna for VHF/UHF Work", a custom Moxon-Turnstile antenna was designed for operation at 54 MHz using a Moxon Ratio Table and 4nec2 antenna modeling software.

• The antenna with be deliberately de-tuned to improve its propagation pattern (making it more hemispheric) by adjusting its dimensions off of the Moxon Ratios.  These changes in dimensions are reflected in the CAD.