Soon after the collapses. (source)


Red Hot Debris. The removal of debris from the collapsed areas requires the safe lifting and maneuvering of very heavy steel beams, often twisted and tangled from the force of the collapse. Some beams pulled from the wreckage are still red hot more than 7 weeks after the attack, and it is suspected that temperatures beneath the debris pile are well in excess of 1,000°F. (source)

John Gross of NIST giving a lecture at the University of Texas


SPOT satellite image of Manhattan, using the HRVIR (High Resolution Visible Infra Red instrument) on SPOT 4, acquired on September 11 at 11:55 AM EST, 3 hours after two planes crashed into the World Trade Center. The colors result from the use of infrared bands to identify the actual fire hot spots (see red spots near the base of the smoke plume). The SPOT satellites orbit at an altitude of 822 km. (c) CNES/SPOT Image 2001 (full_image)
Resolution: 20 meters; Acquisition time: 09/11/01 11:55 AM EDT; Orbital altitude: 822 km
Band combination: Not specified, but presumed to be:
R 0.79 - 0.89um
G 0.61 - 0.68um
B 0.50 - 0.59um
(source)

As part of the World Trade Center disaster response, the NASA Airborne Visible/Infrared Imaging Spectrometer (AVIRIS) was flown over the site on the 16th and 18th of September 2001. AVIRIS measures the solar reflected spectrum from 370 to 2500-nm at 10-nm sampling. For this flight the data were acquired at 1.5-m spatial sampling with image coverage of the entire disaster site. AVIRIS measurements are spectrally, radiometrically, spatially calibrated in the laboratory and validated in flight. Rapid examination of the World Trade Center AVIRIS data in the 2300 nm spectral region showed numerous high radiance targets indicative of burning fires. A new spectroscopic algorithm was implemented to simultaneously solve for the temperature and fractional area of the fires. This algorithm uses the Planck function in conjunction with the full spectral shape measured by AVIRIS to determine the temperature and fractional area of the fire. This spectral algorithm overcomes the ambiguity between temperature and area that exists in single-spectral-band temperature estimation methods. With these AVIRIS data set and new algorithm, 8 hot spot zones were identified in the September 16th data with temperatures ranging from 700K to 1019K and fractional areas from 1.1 to 18%. Analysis of the data set acquired on September 18th showed 7 of the hot spot zones still present with temperatures ranging from 471K to 952K and fractional areas from 0.5 to 36%. These imaging spectrometer derived physical parameters of fire temperature and fractional-area were found useful to the personnel making decisions on the ground. The complete set measurements, analyses, and results of this effort are reported in this paper. (source)