moisture measurement is time-consuming, labor-intensive, and not continuous; hydrologists would prefer "rapid, reliable, economical, and continuous" methods (1:69). In the 1950's, researchers found a correlation between soil moisture a foot deep or more and the scattering and speed of "fast neutrons" of hydrogen; hence the neutron probe (still in use) was invented (16:66-67). While that tool is useful for monitoring in real time, it is still labor-intensive, it requires gravimetric calibration data, and measurements are neither wide-spread nor continuous. Amer et al. reported developments of two electronic sensors (an "electrical resistance sensor" and a "time domain reflectometer") that could be placed in the field and read remotely, even continuously (1:69). The ERS system did work fairly well, and its outputs were continuous; but the failure rate was high; and if the wire leads from the sensor passed vertically from the sensor up through the soil surface--a most logical expectation, water ran down the wires and gave a false reading of "moisture" content (1:82).
A method of learning soil moisture at numerous points over repeated cycles of wetting and drying seems a minimum requirement for accurate (atmospheric climate) General Circulation Models (GCMs) and (hydrologic) Land Surface Models (LSMs), which NASA has explored since 1988 (13:1247). Satellite measurement of land features are obvious as the long-awaited remotely-sensed, virtually automatic way to estimate continuous SMC.
SOIL MOISTURE INFERENCES FROM REMOTE SENSING
Techniques applied to infer soil moisture from satellites and airplanes have varied from passable to poor. Over a dozen attempts are ranked here from the optimistic to the pessimistic. Elements to spot are: How far above the ground were measurements taken; what was the expectable ground resolution scale; was the method radiometry or (active or passive) microwave radar; what was the wavelength of the ...