Introduction to Remote Sensing

Transcription

1 Introduction to Remote Sensing Trainer Khaled Mashfiq 15 / Nov / 2017 Training Module A2 Session 1 Introduction to the Use of Geospatial Information Technology for Drought Risk Management Phnom Penh, Cambodia November, 2017 In partnership with Contents: Physical Basis of Remote Sensing Satellite System Resolution Image Processing Looking at the Earth from Space!

2 Physical Basis of Remote Sensing What is Remote Sensing? Electromagnetic Spectrum Atmospheric Influence Spectral Reflectance 3 What is Remote Sensing? Remote sensing is the science of obtaining information about objects or areas from a distance, typically from aircraft or satellites Copyright: CNES/ASTRIUM, Digital Globe 4

3 Remote Sensing Made Easy Satellite (Sensor) Processed Information Data Transmission Reflected Energy Sun (Source) Electromagnetic Radiation Ground Station Real World 5 Electromagnetic Radiation Form of energy propagation Perpendicular magnetic and electric fields Characterized by frequency and wavelength Travels at the velocity of light Source: NASA, SEOS, Supermanu 6

4 Electromagnetic Spectrum Source: NASA 7 Sources of Electromagnetic Radiation 8 Remote Sensing relies on the measurement of Electromagnetic (EM) energy The most important source of EM energy is the sun Some sensors detect energy emitted by the Earth itself or provide their own energy Source: NASA

5 Atmospheric Influence Radiation is influenced by atmosphere Scattered, reflected or absorbed by particles in atmosphere Rayleigh scattering (- why sky is blue?) Mie scattering (- why clouds are white?) Source: NASA Earth Observatory 9 Spectral Reflectance Incident/Total Energy Reflected Each object reflects, scatters, transmits and absorbs, specific amount of energy Scattered Absorbed Spectral Reflectance: The amount of reflected energy based on individual material properties Transmitted Energy - Physical and chemical state - Surface roughness - Geometric circumstances (Incidence angle)

6 Understanding Spectral Reflectance 60 Spectral Reflectance of Earth Surface 50 Different surface cover exhibits different amount of reflectance in different wavelength of electro magnetic spectrum Percent Reflectance Wavelength (μm) Reflectance Reflected Energy Incident Energy Muck Soil Vegetation Clear Water Turbid Water Slty Clayey Soil Satellite Systems Satellite Types of Satellite Orbits Active vs Passive Sensors Earth Observation Satellites

7 Satellite Man made objects that are in specific orbit around a planet (Earth, in our case) Carries one or more onboard instruments (sensors) to collect data Each satellite is designed in a unique way for a specific purpose or application Source: NASA 13 Types of Satellite Orbit 14 Low Earth Orbit Pole to pole At an altitude of km Sun-synchronous Example: Earth Observation Satellites Medium Earth Orbit Altitude of around 20,000km Semi-Sun Synchronous Example: GPS High Earth Orbit 36000km and moves at the velocity of earth`s rotational speed Geo Synchronous Example: Weather satellites Source:

8 Types of Sensors 15 Passive Sensor ((Depends on sun`s energy) Active (Has its own radiation)

9 Earth Observation Satellites 17 Mostly in low earth, nearly polar and sun synchronous orbit ( km) Carries active or passive sensors Used for weather, environmental monitoring, mapping, etc. Source: NASA 18 Resolution Spatial Spectral Radiometric Temporal

10 Structure of Remote Sensing Data 19 Each image consists of one or more bands in unique wavelength range Spatial Resolution Size of the pixel (edge length expressed in meters) MODIS (250m) vs Landsat 8 OLI (30m) Lesser Cell Size = Higher Spatial Resolution 20

11 Spectral Resolution 21 Number of unique spectral bands and their width Multispectral (3-8 bands) Visible (R,B,G), IR (Near, Mid, Short Wave, Far (thermal) Panchromatic (Single Band) Hyperspectral (several hundred bands) Narrower Bandwidth = Higher Spectral Resolution Radiometric Resolution How well difference in brightness can be perceived (Number of grey values in a band expressed in bits) Gray Scale 4 Gray Scale 16 Gray Scale 256 Better identification of similar objects = Higher Radiometric Resolution Source: SEOS 22

12 Temporal Resolution Time interval between two identical flights (Revisit time) Earth Observing Satellite (14-16 days) Meteorological Satellites (few hours to 1 or 2 days) Useful in multitemporal analysis such as changes in vegetation, growth of cities over time, deforestationafforestation, etc. Continuous data availability = Higher Temporal Resolution 23 Discussion Corner! It will never be perfect! Can you identify any possible trade-offs? Spatial vs Spectral - What happens if both are high? Will data storage be an issue if the resolution is higher? If yes, why? Can we have all four resolutions spatial, spectral, radiometric and temporal, as high as possible? 24

13 25 Image Processing Corrections Image Enhancement Image Classification Image Corrections Source: John R Jensen

14 Image Enhancement: Radiometric Original DN values (30-90) Stretched DN values (0-255) RADIOMETRIC ENHANCEMENT IN ARCGIS NONE STANDARD-DEV = 2.0 HISTOGRAMEQUALIZE MINIMUM MAXIMUM

15 29 RADIOMETRIC ENHANCEMENT IN ARCGIS NONE PERCENT CLIP 2 % ESRI PERCENT CLIP 2 % DRA Image Enhancement: Spatial Encompasses another set of digital processing functions which are used to enhance the appearance of an image

16 Band Composites? 31 Combining three bands (blue, green and red) RGB Spectral Reflectance Curve 60 Spectral Reflectance of Earth Surface 50 Percent Reflectance BAND 2 BAND 3 BAND 4 BAND 5 BAND 6 BAND 7 Muck Soil Vegetation Clear Water Turbid Water Slty Clayey Soil Wavelength (μm) LANDSAT 8 OLI SENSOR BANDS 32

17 UNDERSTANDING COLORS IN BAND COMBINATIONS WHAT WILL BE THE COLOR OF VEGETATION IN 5,4,3(RGB) OF LANDSAT 8 47 Band 2 Band 3 Band 4 Band Red = 47 Green = 10 Blue = 13 MEDIUM TO BRIGHT RED 33 Image Processing : Band Combinations

18 True Color Composite 35 Band Combinations Color IR (Vegetation) 36

19 Thermal Infra-Red 37 Normalized Differential Vegetation Index (NDVI)

20 Play with Band Combination Sentinel True Color Composite Red, Blue, Green 40

21 False Color Composite NIR, Green, Blue 41 Image Classification Supervised Unsupervised Object Based 42

22 Supervised Classification Differentiate classes by using the sample pixels/ training sets selected by the user to represent specific classes Manual process Needs reference areas in the form of Signature File Needs visual interpretation knowledge Requires trained analyst Example: Maximum Likelihood Classification Source: SEOS Create Training Set Develop Signature File Classify Image Unsupervised Classification Uses the statistical distribution of the pixels within feature spaces to differentiate between classes Automatic Pure static/cluster analysis No reference areas identified Example: K means Clustering Source: IDRISI Generate Clusters Classify Image 44

23 Object Based/Oriented Classification Creates objects of different shape and scales based on texture, context and geometry Multi-resolution segmentation Use multiple bands Simultaneous use of different data For example, IR, elevation, shapefiles Sample sites and statistics to classify needs to be identified Best for high resolution data Source: GISGeography Multi-Resolution Segmentation Add Feature Statistics Select Samples Assign Classes 45 Example of UNOSAT Land Cover Classification

24 Discussion Corner 47 Looking at the Earth from Space! Niagara Falls, Canada/United States Copyright: DigitalGlobe Source: NASA Earth Observatory 48

25 Landsat 5 (Thematic Mapper, 30m) Crop Circles, Saudi Arabia Landsat, NASA Earth Observatory Deforestation in Amazon 50

26 Camera on board ISS Dubai s Artificial Islands 51 ISS Digital Camera - The Great Pyramids at Giza, Egypt Source: NASA / ISS 52

27 Landsat 5 (TM, 30m) - Mississippi River Delta, US Discovery mission Lake Natron, Tanzania 54

28 Landsat 7 30m resolution 9-11 World Trade Center 55 View from Space Shuttle Discovery - Grand Canyon

29 Terra (MODIS 250m) Flooding in Bangladesh Source: NASA Earth Observatory 57 Landsat 5 (Thematic Mapper, 30m) Forest Fires in Coahuila, Mexico Source: NASA Earth Observatory 58

30 MERIS - Gulf of Mexico oil spill (2010) 59 Digital Globe (Quick Bird) Srilankan Tsunami 60

31 Aqua (MODIS, 250m) Hurricanes, Pacific Ocean Source: NASA Earth Observatory Aqua (MODIS, 500m) Hubbard Glacier Advancement, Alaska, US Source: NASA Earth Observatory 62

32 Terra (MODIS, 250m) Winter Haze, Bangladesh Source: NASA Earth Observatory 63 Terra (MODIS, 250m) Wolf Volcano, Galapagos Island Source: NASA Earth Observatory 64

33 Landsat 7 (ETM+, Thermal 60m) Urban Heat Island Source: NASA Earth Observatory 65 Landsat 5 30m resolution - Algal Bloom in England 66

34 MODIS 250 m resolution Dust Storm across Red Sea 67 IKONOS (4m) Mayan Ruins, Bajo, Guatemala Source: NASA Earth Observatory 68

35 Kompsat-2 - Namib Desert Applications of Remote Sensing Environmental Monitoring Disaster / Security Monitoring Weather Geology Geophysics Hydrology Volcanology Oceanography Art Archaeology Land cover Topology Mapping and Cartography Visualization General source of information

36 United Nations Institute for Training and Research Institut des Nations Unies pour la Formation et la Recherche Instituto de las Naciones Unidas para Formación Profesional e Investigaciones Учебньıй и научно-исследовательский институт Организации Объединенньıх Наций معھد الا مم المتحدة للتدریب والبحث Palais des Nations 1211 Geneva 10 Switzerland T F Introduction to the Use of Geospatial Information Technology for Drought Risk Management Phnom Penh, Cambodia November, 2017 In partnership with This presentation should not be copied or disseminated in any manner without the express permission of UNOSAT.