Spectrum of asteroid - visual and thermal IR In the optical/IR portion of EMR spectrum, asteroids have 2 components of flux- a reflected solar spectrum in visible and a themal emission component in the IR. The reflected solar EMR always has the same peak wavelength (500 nm), but the IR emission has a peak wavelength that depends on the temperature of teh asteroids surface.
Transmission of Earths atmosphere From the Earths surface, we view the universe primarily through the visible and radio "windows". Much of the thermal IR (a few to a few hundred microns) is blocked by water vapor absorption.
IRAS telescope This is a small (22 inch) telescope that was put into space in the 1980s to observe the sky in wavelengths from about 12 to 100 microns. It far outperformed the largest telescopes on the Earth in this wavelength interval because it was above the absorption (and emission) of the atmosphere. The IRAS did an all-sky survey of IR sources, and the IRAS measurements of asteroid thermal emission are still usfeul.
Asteroid stellar occultation prediction example By combining precise positions of stars and asteroids, you can predict when the shadow of an asteroid (due to its passage in front of a distant star) will sweep across the Earth. If you measure the duration of the shadow passage from one spot, you can, knowing the speed of the shadow across the Earth (from speed of Earth and asteroid) get the linear size of the asteroid (well, from one location you get only the distance across the asteroid along a single line- a chord.
This prediction is for a ~200 km diamter asteroid (19) Fortuna. For an object this size, the central shadow passage would last typically 20 seconds (asteroid moving typically 10 km/sec relative to Earth, shadow 200 km across at widest point). Because of uncertainites in star and asteroid positions, the predictions are not always accurate, but this prediction of a bright asteroid and bright star was pretty close, as seen in next image.
Asteroid stellar occultation results By combining chords measured from telescopes at various places across the shadow path, you can get the size of the projected cross section of the asteroid. This is often fit with an ellipse for simplicity. Here is a fairly well observed event from June 2008 (prediction in previous image) where the asteroid shadow passed over Mexico , southern Texas and Florida.
Because one only needs to measure when and for how long the bright *star* is occulted, one doesn't even need a large enough telescope to detect the asteroid. Thus, one can often do these observations with a small (easily portable) telescope that is equipped with a video camera and a way to measure times precisely. This is easy these days, with GPS time signals.
There is an entire subculture of amateur astronomers who, with no financial compensation, use portable telescopes to observe these events and increase our knowledge of asteroid sizes.
HST image of Ceres Ceres is the largest asteroid (about 950 km across). But even when closest to Earth it is less than 1 arcsec in apparent size. Thus, groundbased telescopes can see little or no detail on Ceres. This is a HST image of Ceres. HST has a resolution of about 0.1 arcsec, but even this is not enough to show much detail.
Earth, Moon and Ceres to scale
How many asteroids are there? The "official tracker of the asteroids" is the Minor Planet Center http://www.cfa.harvard.edu/iau/mpc.html. Astronomers around the world report observations (position on sky and time of observation) of moving objects to the MPC. The MPC first checks if the object is already known. If not, it is given a provisional designation (e.g. 2008 TC3), where the letters are related to when the object was discovered. After enough observations have been made that an object has a good orbit (typically dozens of observations over several years) the object is given a number. As of March 2009, there are over 210,000 numbered asteroids, with a slightly larger number of objects that have provisional designation but are not yet numbered.
The latest estimate is that there are about 1.5 million main belt asteroids larger than 1 km in size. New surveys should be able to see all these 1.5 million, plus millions of somewhat smaller rocks, so that the number of known asteroids will almost certainly exceed several million in the next decade as the new surveys (CSS, PannSTARRS, LSST) get cranked up.