A quadratic time and pointing fit to 26 HD 139139 K2 dips keeps at least 48 percent of its signal after each aperture pixel removal

We test the spatial distribution of the known HD 139139 dip signal in public K2 Campaign 15 data. We fit fixed published event templates with quadratic time and pointing terms. Coverage rules keep 26 events. The full 110-pixel aperture has a combined signal scale of 1.028 relative to published depth marks. We remove each aperture pixel in turn with the same time basis and full-aperture normalization. The smallest remaining scale is 0.498, or 48.4 percent of the full fit. The removed pixel is at stamp row 3, column 7. A different pixel removal gives a residual inside its own shifted-control range. Removing column 7 leaves 0.118, or 11.5 percent, with control standard deviation 0.136. The column test does not give a clear result. Without pointing terms, the minimum pixel residual is negative. We save eight fit settings, 512 shifted schedules per setting, event-group checks, and synthetic pixel-loss checks. Event selection used these same K2 data. Control ranks are descriptive, not calibrated discovery or rejection probabilities. The result measures the fitted signal distribution under the stated baseline model. It does not reject all one-pixel models or exclude a physical fault that affects a full column before calibration. The cause of the dips remains open. The package includes the exact input, provenance, code, tables, and four figures.

HD 139139 · EPIC 249706694 · K2 Campaign 15 · Random Transiter · calibrated pixels · detector systematics · saturation · reproducible photometry