Remote sensing refers to techniques that measure electromagnetic radiation at a distance through “active” or “passive” sensors. Various algorithms are used to convert those measurements into meteorological variables such as surface temperature, vertical profiles of atmospheric temperature, wind and humidity, snow and ice cover, and radiation measurements. “Active” sensors emit electromagnetic waves and measure the portion of waves that scatter back to the sensor from the target. By contrast, “passive” sensors detect radiation that comes to them naturally from other objects, emitted from the landscape/atmosphere or reflected from another source.
These measurements are made by ground-based, airborne, or space-borne instruments which point at targets a considerable distance from their location. For example, ground-based radar instruments emit radio wave energy up through the atmosphere and use measurements of the reflected fraction to estimate the intensity and type of precipitation. Low-level temperature and humidity can be measured from ground-based passive microwave or infrared radiance observations or by active sensors called water vapour differential absorption lidars. The principal remote sensing data we consider in this report are recorded on mobile platforms such as satellites, airplanes, weather balloons, and ships.
Raw data collected by remote sensing must be processed and related to the variable of interest through a "retrieval algorithm". The retrieval algorithm is specific to the instrument and target of observation. Various methods have been developed to remotely sense and retrieve the variables considered in this report, such as temperature, precipitation, humidity (air and soil), snow depth, snow water equivalent, river discharge, sea ice, wind, etc. Snow and ice surfaces are highly reflective to visible light, so a satellite-based sensor measuring in the visible part of the spectrum can distinguish snow- and ice-cover from ocean or other types of land. To detect water in the air (humidity, clouds, or precipitation), infrared or longer wavelengths are typically measured, as in the ground-based radar example above. Most techniques make use of observations at multiple wavelengths to isolate one variable or to extract more information. For example, measurements of microwave radiation at a range of wavelengths can be used to derive vertical profiles of temperature which can be useful for monitoring the climate or individual weather fronts.
On a global scale, the highest volume of observations is provided by instruments on satellites, so special attention is given to them in this section. In particular, remotely sensed data are used to observationally constrain the reanalysis products mentioned in this report, such as the ECMWF reanalysis, version 5 (ERA5). A more thorough introduction to remote sensing from the Canada Centre for Remote Sensing can be found online at https://www.nrcan.gc.ca/sites/www.nrcan.gc.ca/files/earthsciences/pdf/resource/tutor/fundam/pdf/fundamentals_e.pdf.
Earth-Observing Satellites
Satellite data was first collected in earnest in the early 1970s and became a regular part of operational applications from the start of the 'satellite era' in 1979. One major advantage of satellite remote sensing is the capability to quickly observe large regions of the Earth (land and ocean). However, as with any observation-based data, satellite-derived data has issues and challenges. For example, satellite data is limited by the retrieval method. To illustrate this, we use wind as an example. Scatterometers measure backscattered radiation from ocean waves, so surface wind measurements using these instruments are only available over the ocean. We note that these datasets may still be useful for coastal applications in the Canadian North. The remote sensing of winds can also be done via radar and lidar (ground-based or satellite-based, as on the Aeolus satellite), but these active sensors can only measure one component of the winds, called the line-of-sight wind. Finally, yet another wind remote sensing technique tracks the movement of clouds (using atmospheric motion vectors or AMVs) to retrieve winds. However, this method is limited in the vertical direction since clouds are not present everywhere and at every height.
Satellite data is also subject to uncertainties and limitations related to instrument calibration. Other issues that limit remote sensing are spatial coverage, illumination, and sensitivity issues, as well as the presence of the atmosphere (Dubovik et al., 2021).
Clouds cover much of the Earth’s surface at any given time, and they interfere with measurements at visible and infrared wavelengths. However, clouds permit microwaves to pass through, so microwave wavelength measurements are less affected by cloud cover. At high latitudes, passive measurements at visible wavelengths are affected by the lack of sunlight during winter months, whereas active sensors can continue collecting data.
The latitudinal limits and the sampling frequency at each latitude depend on each satellite's orbit. It is therefore crucial to note when “global” datasets have made use of gap-filling techniques beyond the physical limits of the measurement platform. For example, for the snow measurements from the Integrated Multi-satellitE Retrievals for GPM (Global Precipitation Measurement; Tan et al., 2019), "the zone outside 60 N-S is only populated with gauge data [in-situ] and satellite estimates in areas without snowy/icy surfaces, so data coverage is incomplete". Compared to the satellite-derived product, the gap-filled data may have different characteristics, as described in sections 2.1.1 and 2.1.2. It is generally desirable to find a satellite product that covers a region of interest with one measurement technique.
One final issue we outline is "ground clutter", namely the presence of surface conditions that impede the measurement of low-atmosphere or land characteristics. These inherently limit all low-level or surface retrievals over ice, snow, clouds/fog, forests, and complex terrain. This is because remote sensing of the Earth’s surface or lower atmosphere makes use of reflected or emitted signals from the surface and lower atmosphere. “Ground clutter” can interfere with those techniques, yielding incorrect values from the retrieval algorithm. One example of this issue is detailed in Bennartz et al., 2019 in the context of snowfall retrievals over the Greenland Ice Sheets.
Despite these limitations, the data volume and coverage of satellite-derived datasets far exceeds in-situ datasets. Many efforts are underway to combine satellite-derived datasets to extract complementary information. Though the challenges need to be understood, remotely sensed datasets are an indispensable part of weather forecasting and climate monitoring.
Citations
Bennartz, R., Fell, F., Pettersen, C., Shupe, M. D., & Schuettemeyer, D. (2019). Spatial and temporal variability of snowfall over Greenland from CloudSat observations. In Atmospheric Chemistry and Physics (Vol. 19, Issue 12, pp. 8101–8121). Copernicus GmbH. https://doi.org/10.5194/acp-19-8101-2019
Dubovik, O., Schuster, G. L., Xu, F., Hu, Y., Bösch, H., Landgraf, J., & Li, Z. (2021). Grand Challenges in Satellite Remote Sensing. In Frontiers in Remote Sensing (Vol. 2). Frontiers Media SA. https://doi.org/10.3389/frsen.2021.619818
Joe, P., Melo, S., Burrows, W. R., Casati, B., Crawford, R. W., Deghan, A., Gascon, G., Mariani, Z., Milbrandt, J., & Strawbridge, K. (2020). Supersite at Iqaluit: The Canadian Arctic Weather Science Project. In Bulletin of the American Meteorological Society (Vol. 101, Issue 4, pp. 305–312). American Meteorological Society. https://doi.org/10.1175/bams-d-18-0291.a
Mariani, Z., Crawford, R., Casati, B., & Lemay, F. (2020). A Multi-Year Evaluation of Doppler Lidar Wind-Profile Observations in the Arctic. In Remote Sensing (Vol. 12, Issue 2, p. 323). MDPI AG. https://doi.org/10.3390/rs12020323
Moazami, S., & Najafi, M. R. (2021). A comprehensive evaluation of GPM-IMERG V06 and MRMS with hourly ground-based precipitation observations across Canada. Journal of Hydrology, 594, 125929. https://doi.org/10.1016/j.jhydrol.2020.125929
Tan, J., Huffman, G. J., Bolvin, D. T., & Nelkin, E. J. (2019). IMERG V06: Changes to the Morphing Algorithm. Journal of Atmospheric and Oceanic Technology, 36(12), 2471–2482. https://doi.org/10.1175/jtech-d-19-0114.1
Additional e-learning resources:
- https://www.meted.ucar.edu/education_training/
- https://www.ecmwf.int/en/learning/elearning-online-resources