During the work carried out at the KORICANCHA, we conducted topographic surveys of the Santo Domingo Convent, using the Universal Transverse Mercator (UTM) system as the reference, zone 19, with elevations above mean sea level provided by differential leveling performed by the National Geographic Institute (IGN) of Peru in 1970.

We used an Electronic Tachymeter (total station) by LEICA, Model TC-650, with a precision of +/- 5 mm plus 5 ppm in distance. A Distancemeter by WILD, Model DI-4L, integrated into a Theodolite by WILD, Model T2, with a distance precision of +/- 5 mm, was also used.

1. Digital Survey of the Floor Plans of the Church, the First and Second Cloisters of the Convent; First and Second Floors, Roofs, and Ceilings of the Church and the two Cloisters.

2. Digital Survey of the Koricancha – Site Museum, showing all visible details such as Terraces, Pathways, Stairs, etc., with contour lines every 0.50 meters.

3. Digitized Survey of Five Cross-Sections, each with double elevation of pre-established sections using the Total Station and Tachymetry Equipment by insertion (two theodolites equipped with laser eyepieces, angled eyepieces, and a field computer with intersection software).

The subsoil of the Koricancha was explored with a Ground Penetrating Radar (GPR) made by Geophysical Survey Systems Inc. (GSSI). We used an SIR-2 console, with a Motorola DSP 56002 radar processor and a 21 cm VGA LCD screen, connected through a 30 m cable to antennas of 100, 200 and 400 MHz, together with a laptop and data-transfer software.

During the planning of the campaign, a grid was laid out over the whole explorable surface of the Koricancha and the Santo Domingo Convent. Passes were made at an almost constant speed for each antenna, always in the direction marked by the grid. Data were acquired between 17 August and 2 September 2000.

In total, 2,052 profiles were recorded: 1,145 with the 400 MHz antenna, 668 with the 200 MHz antenna and 116 with the 100 MHz antenna. The remaining 123 were rejected during acquisition because they did not meet the quality criteria. Most profiles were recorded in continuous mode; at places of special interest, such as the main altar and the room giving access to the bell tower, static shots were also taken.

The equivalent depth is the depth covered by the recording window; the depth the signal actually reached depends on the ground at each point.

The optimum parameters were calculated for the characteristics of the ground. Gain was first adjusted manually, but because the ground was so heterogeneous it was set to automatic mode to save time: this sped up acquisition without affecting quality, although each profile needed careful checking before it was recorded. The data were filtered only with a low-frequency cut at f/6 and a high-frequency cut at 2f (where f is the antenna frequency), plus a horizontal smoothing of 2 for the 100 MHz antenna. No other filters were applied, so as not to distort the original data; they were used only for display, to remove noise.

The data were processed with GSSI's RADAN software for Windows NT and its 3D QuickDraw extension, which interpolates a series of parallel profiles to build a three-dimensional volume that is easier to interpret. Each profile was first filtered to remove noise and then adjusted horizontally, so that the distance between marks is always constant, and vertically, to fit the topography of the ground.

The radar recorded structures, crypts and cavities of colonial, Inca and pre-Inca date beneath the church. In 2000, however, it did not locate the crypt with the entrance to the tunnel (Chinkana) through which Anselm Pi Rambla, Francesc Serrat and the then Prior of the convent entered in 1982. In 2020, a remote-sensing specialist in Spain re-analysed the 2000 profiles with state-of-the-art software and found the reason: the radar had been programmed with a compressed depth, which is why it never found the crypt. In 2023, with a far more advanced ground-penetrating radar, the Foundation located the exit of the tunnel.