Question 1 of 5: Remote Sensors and Their Characteristics

Question 1: Remote Sensors and Their Characteristics (20 marks — 5 each)

Question text not reproduced: the examination questions are © Engineers and Geoscientists BC. Open the official past paper (linked at the top of this page) to read the question, then follow the worked solution below.

1.1 Pushbroom imager versus optical-mechanical (whiskbroom) scanner

An optical-mechanical (whiskbroom) scanner sweeps a single detector — or a small detector bank — across the swath with an oscillating or rotating mirror, building each scan line one pixel at a time. A pushbroom imager instead uses a fixed linear array of detectors, one per across-track pixel, and forms the image line-by-line as the platform's forward motion "pushes" the array along the ground track; no scanning mirror is required.

Advantages of the pushbroom. Because every ground cell in a line is viewed by its own dedicated detector, the dwell (integration) time per pixel is far longer than the microseconds a whiskbroom detector spends on each cell. The longer dwell collects more photons, which raises the signal-to-noise ratio and thereby permits finer spatial resolution, finer radiometric resolution, or narrower spectral bands for the same energy budget. The absence of a moving scan mirror also makes the instrument lighter, mechanically simpler, more reliable, and less power-hungry, and it fixes the interior geometry of a line (no mirror-velocity distortion), giving more stable geometric fidelity.

Disadvantages of the pushbroom. A line now contains thousands of individual detectors that can never be perfectly identical; small differences in their gain and offset produce coherent along-track striping that must be removed by careful, ongoing relative radiometric calibration. Manufacturing long, uniform arrays that are also sensitive far into the short-wave-infrared or thermal region is difficult and costly, so pushbroom designs historically covered fewer bands than a whiskbroom that reuses one detector for the whole swath. Any single dead or drifting detector corrupts an entire image column rather than scattered pixels.

1.2 Why passive thermal and microwave sensors have low spatial resolution

Two independent effects push the spatial resolution of passive thermal-infrared and passive-microwave sensors toward coarse ground-cell sizes. First, the available radiant energy is weak. Emitted terrestrial radiation at thermal (≈8–14 µm) and especially at microwave (millimetre-to-centimetre) wavelengths carries far less energy per unit bandwidth than reflected solar radiation in the visible/near-infrared. To collect enough energy for an acceptable signal-to-noise ratio, the instantaneous field of view (IFOV) must be enlarged so that each detector integrates radiation from a bigger patch of ground — directly coarsening the spatial resolution.

Second, the diffraction limit scales with wavelength. The finest angular detail an aperture of diameter $D$ can resolve is of order $\theta \approx 1.22\,\lambda/D$; for a fixed, launch-realistic aperture the resolvable angle grows in proportion to $\lambda$. Thermal wavelengths are roughly twenty times longer than visible light, and passive-microwave wavelengths are thousands of times longer, so the diffraction-limited ground spot is correspondingly large. Both the energy argument and the diffraction argument therefore force long-wavelength passive sensors toward low spatial resolution.

1.3 Three sensor characteristics other than spatial resolution

Spectral resolution — the number, width, and placement of the wavelength bands a sensor records. A multispectral sensor with a few broad bands has coarse spectral resolution; a hyperspectral sensor with hundreds of contiguous narrow bands has fine spectral resolution and can resolve subtle absorption features.

Radiometric resolution — the sensitivity of the sensor to differences in radiance, i.e. the number of quantization (brightness) levels used to record the signal, usually expressed in bits. An 8-bit sensor stores 256 levels; a 12-bit sensor stores 4096, and so discriminates finer differences in reflected or emitted energy.

Temporal resolution — the revisit interval, i.e. how frequently the sensor can image the same ground location (for example, 16 days for Landsat, or daily for a wide-swath system). It governs how well change and seasonal dynamics can be monitored.

(Swath width — the across-track ground extent imaged in one pass — is an acceptable fourth characteristic and trades off against spatial resolution.)

1.4 Active versus passive sensors

A passive sensor records energy that originates from an external source: reflected sunlight in the visible/near-infrared, or radiation thermally emitted by the scene itself. It carries no illumination source of its own, so its solar-reflective bands work only in daylight and its measurements depend on uncontrolled illumination geometry and atmospheric state. Examples are Landsat ETM+, SPOT, and MODIS. An active sensor supplies its own energy, transmits a controlled pulse toward the target, and measures the returned signal — radar (SAR), lidar, and scatterometers are examples.

The contrast follows directly. An active sensor controls the wavelength, polarization, timing, and geometry of the illumination, so it can operate day or night and (at radar wavelengths) see through cloud and haze, and it can measure range/height precisely from the two-way travel time. That capability costs transmitted power, heavier hardware, and more complex signal processing (speckle, phase). A passive optical sensor is simpler and lighter and directly senses the target's natural spectral reflectance/emittance, but it is hostage to solar illumination, shadows, and clouds. In short, the two are complementary: passive systems characterize natural spectral response, while active systems provide controlled, all-weather, geometry-rich measurements.