AFDØS enhanced explorers: data and methods
AFDØS examines reported frequency variations from receiving stations monitoring M7SQI. Transmitting frequency is assumed reasonably stable over an operating interval but has not been independently verified. Variations may reflect propagation, transmitter or receiver frequency errors, calibration effects and changing receiver populations. These plots do not establish ionospheric Doppler effects or travelling disturbances.
Selection and analysis
Choose a period, region or map corridor. Additional summaries use the actual selected observations. Link time axes is enabled initially; visible-interval recalculation is a separate option, initially disabled. Histograms are not time axes. Each graph has an independent disclosure and vertical scale. Reset restores the full selected interval. Spectra run only when explicitly requested and become invalid when the analytical selection changes.
Nominal-frequency comparison
The FTx original 15-second comparison and nominal TX reference remain separate from calibrated residuals. Raw observations without an eligible baseline are shown separately; no corrected frequency is invented. Baseline eligibility requires three records in the existing monthly receiver/source/band/mode group, not three records for each transmission. Zero timing-matched stored observations is not proof of no reception, and missing TX logs are not proof of inactivity.
Geography
The original draggable SVG maps and path filters are retained. FTx computes distances from the centre of IO70pq (50.6875°, −4.7083333°). WSPR retains its original transmitter coordinate rounding, locator precision and reported-distance filter. Miles are a display conversion of the original kilometre thresholds. WSPR frequency populations remain separate; overlapping month, run, day and week selections are alternate views, not independent observations.
Detailed reporting
The explorer offers selection summaries, technical explanations and a diagnostic JSON download. A future HTML report for an individual period should include source generation and checksum, UTC selection, geometry, baseline eligibility, receiver composition, gap and spectrum settings and an explicit separation of descriptive results from physical interpretation. Thousands of static period reports have not been generated.
Spectral panels and residuals
Spectra and timestamp sampling windows run only when you click Analyse current selection beside a panel. Each panel explains pending or unavailable results. More than2000 bins deliberately withholds frequency estimation; timestamp windows are independent of that guard and frequency variance. Gaps, a constant detrended series, too few bins or an invalid period range have distinct explanations. A maximum gives descriptive fitted period/amplitude and fit improvement, not validated significance. The new residual panel subtracts the existing local median reference only for display; it does not correct the observations or feed the spectral fit. Legends sit below each graph.
Period display and plot explanations
Each frequency-spectrum panel has an independent Automatic / Minutes / Hours selector. Short spectra default to minutes; longer WSPR spectra default to hours. Switching converts the cached x-axis, hover values, peak marker and summaries without rerunning a fit. Machine-readable period_min fields remain in minutes; the separate sampling-window graph labels its shared axis in minutes. Numeric maxima are fitted candidates, not confirmed physical oscillations. Legends wrap below plots; narrow panels may grow taller to keep labels readable. Each plot has visible measurement/purpose notes and expandable method/limitation details.
Growing archive and data snapshots
The operator manually prepares and publishes a consistent snapshot from existing reception archives and TX logs. A page keeps its published data version while open; reload to see a newly published snapshot. The historical selector retains the exact Stage2B reference on these same explorer pages. Preparation can revise monthly FTx receiver/source/band/mode corrections and WSPR receiver frequency-population medians. No calibration is automatically frozen when a TX session ends. A published version stays unchanged; revisions appear only in a new version.
The status dates describe the whole snapshot, not just the selected plot. Imported reports may be newer than calibrated plotted reports because matching or baseline eligibility is unavailable. Sparse WSPR populations with fewer than ten displayed reports may require lowering the existing Minimum reports filter. No gaps are filled and no transmission is invented from reception data.
Review status
This is a Stage2C development candidate extending the established enhanced explorers. Local PHP and in-app browser checks have been performed. Firefox, touch-device interactions and actual shared-host configuration still need review. No deployment has occurred.