ENVS211W - Introduction to Remote Sensing PDF

ENVS211W Introduction to Remote Sensing 1  Why use remote sensing? ▪ It is an important source of information for a variety of features: oceans, atmosphere, ice, weather, agricultural and natural resources, monitoring disasters ▪ Monitor and develop under...

ENVS211W Introduction to Remote Sensing 1  Why use remote sensing? ▪ It is an important source of information for a variety of features: oceans, atmosphere, ice, weather, agricultural and natural resources, monitoring disasters ▪ Monitor and develop understanding of environment ▪ Information can be accurate, timely, consistent and cover small to large areas ▪ Some historical data (from 1970s until now) Advantages ▪ Cheapest way to repeatedly view the entire Earth ▪ Digital data (easy to manipulate) Disadvantages ▪ High initial cost (100-500 million dollars to build and launch) ▪ High-tech 2  What is remote sensing It is the science and art of acquiring information about the Earth's features without actually being in contact with them This is done by sensing and recording reflected or emitted energy and processing, analyzing, and applying that information “Everything in nature has its own unique distribution of reflected, emitted, and absorbed radiation. These spectral characteristics can – if ingeniously exploited – be used to distinguish one thing from another or to obtain information about shape, size, and other physical and chemical properties” – Parker & Wolf, 1965 3  Elements of Remote Sensing 1. Energy Source or Illumination (A) – the first requirement for remote sensing is to have an energy source which illuminates or provides electromagnetic energy to the target of interest. 2. Radiation and the Atmosphere (B) – as the energy travels from its source to the target, it will come in contact with and interact with the atmosphere it passes through. This interaction may take place a second time as the energy travels from the target to the sensor. 3. Interaction with the Target (C) - once the energy makes its way to the target through the atmosphere, it interacts with the target depending on the properties of both the target and the radiation. 4. Recording of Energy by the Sensor (D) - after the energy 5. Transmission, Reception, and Processing (E) - the has been scattered by, or energy recorded by the sensor has emitted from the target, we require a sensor (remote - not in to be transmitted, often in electronic form, to a receiving and contact with the target) to collect processing station where the and record the electromagnetic radiation. data are processed into an image (hardcopy and/or digital). 7. Application (G) - the final element of the remote sensing process is achieved when we 6. Interpretation and Analysis (F) - the processed image is apply the information we have been able to extract from the interpreted, visually and/or imagery about the target in order digitally or electronically, to extract information about the to better understand it, reveal some new information, or target which was illuminated. assist in solving a particular problem. 4  Satellite Orbits 1. Geostationary Orbit: Circular orbit 35,786km above surface Orbits around equator It rotate with the same speed of the Earth, so it appears to be in one spot Commonly used for: – Communications – Weather satellites – See example: https://eumetview.eumetsat.int/static- images/MSG/RGB/NATURALCOLOR/SOUTHERNAFRICA 5 /index.htm  Satellite Orbits 2. Polar Orbit: It collect data from the earth during the daytime because it captures data when the Sun is overhead Usually sun-synchronous, i.e. crosses the equator at the same time every day Commonly used for: – Communications – Earth Observation 6  Satellite Orbits Polar Orbit: 7  Types of remote sensing sensors Passive or Optical Active The sensor records energy that is The sensor illuminates radiation to the reflected or emitted from the source, object and record it back, such as such as light from the sun. radar or microwaves 8 Optical /Passive remote sensing 9  Optical /Passive remote sensing Limitations of optical ▪ Cloud cover is the main limitation of optical/passive remote sensing 10  Active / radar / microwave remote sensing Active sensors provide their own energy source for illumination. The sensor emits radiation towards an target record it back. Advantages for active sensors include the ability to obtain measurements anytime, regardless of the climatic conditions 11 Physical basis of remote sensing An understanding of electromagnetic energy and its properties is important in order to interpret remotely sensed data Electromagnetic energy: energy propagated through space or through material media in the form of an interaction between electric and magnetic fields, moving at the velocity of light Forms of electromagnetic energy: visible light, radio waves, heat, ultraviolet rays, and x-rays 12 Electromagnetic Radiation 13  Electromagnetic spectrum Essentially a continuum of energy waves of the same speed but different wavelengths and frequency The visible region (blue, green, red) is defined by the sensitivity to the human eye; stretching from ~0.4-0.7 μm The main regions used in RS are visible, infrared (~0.7-3 μm) and microwave (~1mm-1m) The thermal region (~3-14 μm) is being explored! Consequently, several regions of the electromagnetic spectrum useful for remote sensing… 14  Electromagnetic spectrum Most sensors record within the visible (V), near infrared (NIR) and within the shortwave infrared (SWIR). 15  Absorption Absorption is the process by which radiant energy is absorbed and converted into other forms of energy. An absorption band is a range of wavelengths (or frequencies) in the electromagnetic spectrum within which radiant energy is absorbed by substances such as water (H2O), carbon dioxide (CO2), oxygen (O2), ozone (O3), and nitrous oxide (N2O). 16  Reflection Reflection (R) occurs when radiation bounces off the target and is redirected 17  Scattering Scattering occurs when particles or large gas molecules present in the atmosphere interact with and cause the electromagnetic radiation to be redirected from its original path. Rayleigh scattering occurs when particles are very small compared to the wavelength of the radiation. These could be particles such as small specks of dust or nitrogen and oxygen molecules. Mie scattering occurs when the particles are just about the same size as the wavelength of the radiation. Dust, pollen, smoke and water vapour are common causes of Mie scattering which tends to affect longer wavelengths 18 Electronic Imagery ▪ Image composed of pixels ▪ Record photons reflected/emitted energy ▪ Usually square ▪ Every cell has a value ▪ Represents brightness value 19  Digital data ▪ One value per pixel ▪ Digital value recorded as series of binary values bits ▪ Size of value depends on the storage space that is allocated per pixel 20  Digital representation ▪ Storage space is measured in binary numbers ▪ One byte can store 8 bits ▪ 11111111 = 1x27+1x26+1x25+1x24+1x23+1x22+1x21+1x20 = 128 + 64 + 32 + 16 + 8 + 4 + 2 + 1 = 255 ▪ Therefore the largest value that can be stored in an image with 1 byte storage space (radiometric resolution) per pixel is 255 ▪ A Landsat image has a radiometric resolution of 1 byte or 8 bits (256 values) ▪ 16 bits? ▪ 216 = 65536 21 The human concept of colour… 22  Sensor Platforms Three basic levels: 1. Spaceborne – Space vehicles (satellites) launched into space – Very expensive – Low (kilometer) to high (decimeter) resolutions – Each satellite carries a payload of sensors & instruments Satellite & sensors must: – Stay in stable orbit – Generate power (from solar panels) – Capture information – Store information – Relay information back to earth 23  Sensor Platforms 2. Airborne – From camera mounted on an aeroplane – Space vehicles (satellites) launched into space – Usually within a planned campaign with a pre-determined flight path – Centimeter resolution 24  Sensor Platforms 3. Terrestrial – Most likely not universally accepted as “remote sensing” – Includes terrestrial photogrammetry and scanning LIDARsFrom camera mounted on an aeroplane 25  Visual Image Interpretation We make sense of what we see by interpreting a large variety of elements in visual scenes Remote sensing imagery has increased challenges to the interpreter because of – An unfamiliar (aerial) perspective – The use of wavelengths outside the visual spectrum – The use of unfamiliar scales & resolutions 26  Visual Image Interpretation Elements of image interpretation: Shape: this characteristic alone may serve to identify many objects. Size: noting the relative and absolute sizes of objects is important in their identification Image Tone or Color: all objects reflect or emit specific signatures of electromagnetic radiation Pattern: many objects arrange themselves in typical patterns. Shadow: shadows can sometimes be used to get a different view of an object. Texture: imaged objects display some degree of coarseness or smoothness. 27  Sensor Characteristics The ability of the sensor to identify features depends on 4 types of resolutions: ▪ Spatial resolution ▪ Spatial resolution of the sensor refers to the size of the smallest possible feature that can be detected ▪ Spatial resolution is limited by the pixel size, i.e. the smallest detectable object cannot be smaller than the pixel size ▪ High resolution imagery have small pixel size and smaller features can be recognized clearly – suitable over smaller areas ▪ Low resolution imagery have large pixel size and larger features can recognized – suitable over broader areas 28 Sensor Characteristics 29  Sensor Characteristics ▪ Spectral resolution ▪ The number, wavelength position and width of spectral bands a sensor has ▪ A band is a region of the EMR to which a set of detectors are sensitive ▪ Multispectral sensors have a few, wide bands (several spectral bands) ▪ Hyperspectral sensors have a lot of narrow bands (hundreds of spectral bands) 30  Sensor Characteristics ▪ Temporal resolution ▪ The time needed to revisit and acquire data for the exact same location ▪ The higher the temporal resolution, the shorter the interval between the acquisitions of images ▪ Some features change more often, and higher temporal resolution is required to understand the dynamics ▪ Higher temporal resolution may also help in areas affected by frequent cloud cover, because there is high probability to acquire cloudless images 31  What can be measured from Remotely Sensed Imagery? Surface Meteorology – height – pressure – land use activities – temperature – vegetation type – cloud cover – surface (water) – cloud top height – temperature – cloud type – fires – lightning frequency – surface roughness Chemical constitution of – water turbidity / chlorophyll the atmosphere concentrations – aerosol type – ice cover – trace species 32  Remotely sensed indicators Indicators are measurable entities that relate to a condition, change of quality and state of phenomena under investigation. Indicators are useful to monitor changes & provide means to compare trends and progress over time. Indicators must be sufficiently representative & easy to understand & measure on a routine basis Without good quality indicators & relevant data &, the assessment loses: – the ability to reflect the condition of a resource – the ability to measure progress towards sustainability goals & objectives 33 Vegetation Spectral Reflectance Curves MC = moisture content ▪ Orange curve is for stressed plant/vegetation ▪ Green curve is for healthy plant ▪ So, healthy plants reflects highly in the NIR ▪ Stressed plants reflects less in NIR ▪ Healthy plants absorbs well in the SWIR regions 34 ▪ Stressed plants absorb less in the SWIR Spectral Indices https://encrypted-tbn3.gstatic.com/images?q=tbn:ANd9GcTJHBiUxi2tdFIowHv7CenxjHUl79103coSAKIeaT_TVuz70EoJRA 35  Spectral Indices Considerations in the use of NDVI Saturates over dense vegetation Less information than original data Any factor that unevenly influences the red and NIR reflectance will influence the NDVI – such as atmospheric path radiance, soil wetness Pixel-scale values may not represent plant-scale processes Derivatives of NDVI (LAI) are not physical quantities and should be used with caution Disturbance/changes may be caused by different 36 factors Spectral Indices 37 Satellite sensors and their properties Sentinel Landsat MODIS ▪ Images since 1984 ▪ Images since 2014 ▪ Images since 2000 ▪ 30-m resolution ▪ 10-m resolution ▪ 250-m resolution ▪ 16-day revisit cycle ▪ 12-day revisit cycle ▪ 1-2 day revisit cycle ▪ Free of charge ▪ Free of charge ▪ Free of charge Wordview-2 IKONOS QuickBird ▪ Images since 2009 ▪ Images since 1999 ▪ Images since 2001 ▪ 1.84-m resolution ▪ 0.82-m resolution ▪ 0.6-m resolution ▪ Charge ▪ Charge ▪ Charge 38  Image classification This is the science of turning remotely sensed data into meaningful categories representing surface features or classes Assigning pixels to classes Group similar pixels together Classifications are divided into two: 1. Supervised classification and 2. Unsupervised classification 39  Supervised classification ▪ Analyst define and select samples and classify the image based on chosen samples ▪ Identify similar representative samples of the different surface cover types on the map or via field visits, aerial photos ▪ Spectral information used to ‘train’ the algorithm to recognize spectrally similar areas for each class ▪ Algorithm (several available) used to determine numerical ‘signatures’ for each training class ▪ Each pixel is compared to class signatures and labeled as the class it most closely ‘resembles’ digitally 40  Supervised classification ▪ Often requires ancillary data (maps, photos, etc.) ▪ Field work often needed to verify ▪ Have several training areas for one category ▪ Locations must be spread around the image and be easily transferred from map to image 41  Supervised classification Advantages ▪ Analyst has control over the selected classes tailored to the purpose ▪ Has specific classes of known identity ▪ Does not have to match spectral categories on the final map with informational categories of interest ▪ Can detect serious errors in classification if training areas are misclassified Disadvantages ▪ Analyst imposes a classification (may not be natural) ▪ Training data are usually tied to informational categories and not spectral properties ▪ Training data selected may not be representative ▪ Selection of training data may be time consuming and expensive ▪ May not be able to recognize special or unique categories because they are not known or small 42  Unsupervised Classification Advantages ▪ Requires no prior knowledge of the region ▪ Human error is minimized ▪ Unique classes are recognized as distinct units Disadvantages ▪ Classes do not necessarily match informational categories of interest ▪ Limited control of classes and identities ▪ Spectral properties of classes can change with time 43  Accuracy assessment Is the measure of how well a classification worked Collect reference data: “ground truth” Determination of class types at specific locations – Compare reference to classified map Does class type on classified map = class type determined from reference data? Accuracy Assessment: Reference Data – Aerial photohraphs – High resolution satellite imagery – Ground truth with GPS – GIS layers 44  Accuracy assessment Number of correctly classified site: 21 + 31+ 22 = 74 Total number of reference sites = 95 45 Accuracy assessment 46