Lithospheric correction sets the error budget of satellite equatorial electrojet retrievals over the Bangui anomaly

The Bangui magnetic anomaly lies beside the dip equator in central Africa, and its crustal field enters every satellite measurement of the equatorial electrojet made at that longitude. We quantify that contamination and test whether the anomaly alters the storm-time electrojet. Scalar magnetic data from Swarm A and B were analyzed for the St. Patrick's Day storm of March 2015 and the Gannon storm of May 2024, together with the five international quietest days of each month. A line-current inversion with fixed regularisation returned the peak sheet current density for 130 dayside equator crossings in four longitude sectors. Three lithospheric models were removed in turn, and the retrieval was repeated with no crustal correction. Over Bangui the median crustal signal inside the electrojet core reaches 4.45 nT against an electrojet signal of 16.80 nT, and removing it changes the retrieved peak current by 45 % at the median, against 7 % to 17 % at three control longitudes. The three models differ by 0.58 % at Bangui, so the correction is tightly constrained where it is largest. Inside the current layer at 108 km the lithospheric field reaches 0.6 % of the core field at most and displaces the dip equator by 2.3 km, which makes the contamination instrumental rather than physical. Retrieved intensities track the Huancayo ground magnetometer (ρ = 0.91) and the ESA equatorial electric field product (ρ = 0.49). Eastward intensities over Bangui match the control sectors in the quiet and recovery phases, with p ≥ 0.29 throughout, and no lagged response to the interplanetary electric field or to AE survives a family-wise permutation test. Swarm A and B reach the dayside equator between 07 and 12 h local time, which leaves the storm main phase unsampled and limits the power of those comparisons. Short summary Satellites measure the equatorial electrojet, a strong current high above the equator, by sensing its magnetic field. Central Africa holds a huge magnetic anomaly in the rocks below, and we show that this buried rock signal shifts the measured current by about 45 percent, far more than at other longitudes. Crustal field models remove it reliably. Any study of the electrojet over Africa has to correct for the rocks first, or it will mistake geology for ionospheric physics.