William Campbell began his presentation by introducing remote sensing as the analysis of data acquired by a device that is not in contact with the object, area, or phenomena under investigation. He discussed how satellites operate and the types of processing and transformation data undergoes between collection and presentation. Remote sensing instruments are built to look at the electromagnetic spectrum. A passive sensor, such as those on the Landsat satellite, receives data in the form of electromagnetic energy which is translated into an electrical current and then to a digit that represents a pixel (or "picture element") which will then be averaged to form a larger image.
Throughout his talk, Campbell emphasized four keys to successful remote sensing: timeliness, accuracy, resolution, and ground-truthing. (1) Timeliness is important because of the many applications of real time data. For example, the ability to detect hurricanes more accurately could save states millions in evacuation procedures -- Campbell noted that it takes $1 million per acre of coastline for evacuation before a hurricane. Timeliness is also crucial to environmental change detection -- such as the route of an oil spill. (2) In considering accuracy, Campbell commented that the electromagnetic sensors have a very high level of accuracy and are able to compensate for the "noise," such as space dust, that the sensor might also pick up. The "birds" are constantly corrected to assure accuracy. He also mentioned that even military intelligence mission are not devoid of error -- the Bosnian war is an example of many such mistakes in accuracy with the US mistaking cars for tanks and embassy buildings as weapon storehouses. (3) Resolution issues are similar to those of accuracy -- trying to determine the differences between participants at the workshop or even that their was a workshop building would be futile at 1 km resolution -- and sometimes image interpretation is trying to convert "an optical illusion into reality." (4) Campbell also stated that there is no replacement for ground verification because space satellites have been designed to observe large scale global change, not localized change.
Campbell then discussed remote sensing technology in the coming decades. Remote sensing systems are rapidly improving and offering higher and higher resolution data both in the governmental and commercial sectors; active sensors such as Canada's RADARSAT and the US Lidar observation systems have increased the application potential of satellites; costs are generally decreasing, although remain high; and the focus is shifting to providing users with timely and accurate data. Campbell has been developing the potential for handheld devices, which can be used by an individual in the field to make in situ observations to verify the satellite data. The ground observer will be able to immediately send the information gained on the ground to a data processing center where it will be used to verify and correct the images produced from satellite data. A new type of sensor, hyperspectral sensors, have great potential to improve accuracy as well. Hyperspectral sensors separate the electromagenetic channels into small bands of energy, yielding more detailed data. Eventually, the spectral signals of objects will be classified and hyperspectral sensors will be able to identify an object by its signal.