Ge111A Remote Sensing and GIS Lecture Remote Sensing - many different geophysical data sets. We concentrate on : Imagery (optical, infrared and radar) Topography Geographical Information Systems (GIS) a way to organize the imagery as well as point, line, and shapefile data; useful for cataloguing and searching regional data bases Note: Positions and Positions (GPS lecture next week) For more info there are Caltech classes: Introduction to the Physics of Remote Sensing (EE/Ae 157 ab) Geographic Information System for Geological & Planetary Sciences (Ge110)
Why use GIS in a field geophysics class? Understand what is in the field as best you can before you go there: Terrain & topography Geology Roads Access and land ownership (wilderness; military land; etc) Geomorphic features (faults, mountain ranges, drainages) Add your own data and locations to the map (locations of survey points and/or lines) Easily produce base maps showing where different surveys were conducted during the class activities
Equations relating wavelength, frequency, and speed: λ=c/f f=c/λ If the wave travels at the speed of light, c c=0.3 m/ns = 3x10 8 m/s. A wave with a frequency of 10 15 Hz has a wavelength, λ, of 3 x 10-7 m, which is 300 nm or 0.3 µm in the ultraviolet part of the spectrum. Thought questions: 1) What happens to the wave if it travels in a medium with speed less than the speed of light? 2) Can you find the mistake in the graph on this page?
Measurements conducted from: Satellites Aircraft Handheld sensors Character of imagery is based on the reflectance and backscatter characteristics of the surface, f(λ) Different materials have different spectral behavior (rocks of different kinds, water, vegetation ) Both material type + physical state of material (grain size, weathering) are important
Ways you could correct for atmospheric absorption Make atmospheric observations simultaneous with the remote sensing (hard to get usually) Use an atmospheric model of absorption based on other dates or locations Make surface spectrometer measurements for calibration, during the survey or during similar season and time as original survey Don t use bands in the spectral area of max. absorption
Atmospheric absorption Spectra of common vegetation + Spectra of common rocks/minerals
Landsat: Only 7 spectral bands, not very useful for discerning material types But because of large image spatial extent and reasonable resolution, good for overview
ASTER (14 bands) Instrument VNIR SWIR TIR Bands 1-3 4-9 10-14 Spatial resolution 15m 30m 90m Swath width 60km 60km 60km Cross track pointing ±318km(± 24 ) ±116km(± 8.6 ) ±116km(± 8.6 )) Quantization (bits) 8 8 12 Note: Band 3 has nadir and backward telescopes for stereo pairs from a single orbit.
Example: Aster band combination Saline Valley Assign different λ bands or combination of bands to RGB to form color image Thermal infrared bands 13, 12 and 10 as RGB Variations in: quartz content appear as more or less red; carbonate rocks are green mafic volcanic rocks are purple
Sensitive to: energy states of electrons in outer shells of transition metals (visible wavelengths) Twisting, rotation, vibrations of bonds in compounds (3-14 micron region) From Hunt (1977) spectral locations of absorption signals for different minerals and rocks
Critical questions to ask when using imagery 1. Spatial resolution (pixel size) and does this vary for some reason? 2. Image extent (General rule: target is always on the boundary) 3. Wavelengths 4. $$$$ Common systems Platform Pixel (m) Extent (km) Cost ($) Aster 15/30 60 Free Landsat 4,5,7 15/30 180 $400+* SPOT** 5/10 60 O(1000) Ikonos*** 1/4 10 O(1000) Planes/Helicopter O(10cm) 10**** ---- + Quickbird * A variety of cheaper combos exist ** French *** Military **** Camera + height above ground
Hyperspectral Imagery Multiple bands (images) each at different wavelengths e.g. AVIRIS - 224 bands Large data volumes!
What is the advantage of hyperspectral images? Much narrower wavelength bands easier to see smaller features in the absorption spectrum.
At radar wavelengths, the atmosphere is transparent Frequencies and Wavelength of the IEEE Radar Band designation Band Frequency (GHz) Wavelength (cm) L 1-2 30-15 S 2-4 15-7.5 C 4-8 7.5-3.75 X 8-12 3.75-2.50 Ku 12-18 2.5-1.67 K 18-27 1.67-1.11 Ka 27-40 1.11-0.075
SAR/InSAR Platforms Aircraft: Shown here: AIRSAR Measures topography, ocean currents Satellites: Repeat pass Fly over once, repeat days-years later Images Measures deformation and topography Space shuttle: Shuttle Radar Topography Mission (SRTM) From: H. Zebker Both from: JPL
Radar is active imaging Natural image coordinates are in units of time: along track & line-of-sight (LOS) range foreshortening layover shadows Imaging radar is side looking (why?) Achieve resolution by clever combination of consecutive radar images: Synthetic Aperture Radar (SAR)
Topography (DEM, DTM, DTED, topo, height, ) Methods Land surveys (now GPS or total station) Radar altimeter Air or space borne laser - point or swath mapping altimeter Stereo imagery (air photos, also now satellite) Radar interferometry a.k.a. InSAR (plane, shuttle, satellite) Optical interferometry a.k.a. LiDAR Practical availability U.S.: 10-30 m/px (USGS, SRTM) on the net 0.5-15 m (Airborne InSAR, optical, laser swath) - e.g., TOPSAR Foreign: 90 m/px (SRTM 60S-60N), 30-60 m/px by begging (classified) 900 m/px open access Make your own (InSAR, optical) 10-20 m/px
Practical Concerns with Imagery and DEMs 1. Continuity of adjacent images 2. Reference mapping information Origin Georeferencing how many tie points are needed? Datum (WGS84, NAD27, NAD83) Projections UTM - eastings and northings (m) Geographic - longitude and latitude (deg) 3. File format # px in x and y coordinates How to store multiple bands (BIL, BIP) Precision (bytes/band/pixel) - always in binary 4. Software (raster + vector) ESRI - ArcGIS ERDAS - Imagine Matlab/IDL (ENVI software) 5. Imaging combinations Shaded relief (intensity) + color (something else) Use Google Earth for simple tasks
The next few images are from Jane Dmochowski s PhD thesis (Caltech Seismo Lab, 2005) Isla San Luis is an active volcanic island in the Gulf of California (Mexico) The imagery is Modis-Aster Simulator (MASTER) airborne data, with about a 4 m pixel size. It was collected with a very lowflying small airplane. The MASTER sensor has 50 spectral bands from visible to thermal infrared (TIR).
LIDAR light detection and ranging works at optical frequencies; see Fugro s LiDAR Fact Sheet LIDAR images of San Andreas fault from P4 project (high resolution topography) can detect the ground below the trees (multiple return LiDAR processing)
Cajon Pass I-15 Fault Crossing
Another example of LIDAR data for topography along the San Andreas fault
Ge111a GIS project (status as of today) Topo maps 1:24k and 1:100k (USGS) Imagery - NAIP, ASTER, Landsat DEM - NAIP (National Agricultural Imaging Project) Hillshade - NAIP DOQQ - 24K and 100K Geographic features (roads, rivers) Township/range/section grids (BLM) Regional land status data base Data base of Quaternary faults (USGS)
Homework part 2 due Thurs March 5 th, 2009 1. Using the class GIS project, construct a basemap(s) of the NE Salton Sea region including the Mecca Hills and the Orocopia Mtns/Diligencia Basin, showing the mapped Quaternary faults. Annotate your map with any geologically or culturally important features (other faults, major alluvial fans, place names, names of faults, etc.) and include scale bars, a geographic reference (latitude/longitude ticks), and a North arrow, as well as legends for any colors or symbols that you use. Print out your map to turn in, but save the file because you will use it later on in the class. 2. Make a perspective image of the Mecca Hills using Google Earth or similar product (based on aerial photographs and an unknown DEM). Sketch on it the locations of major faults. Turn this in with your homework. 3. Use the two maps/images above as well as the results of Part 1 of this homework (the part you did using the GeoCommunicator web site). Write a page answering these two topics: What features are offset by the San Andreas fault in this area? How could you tell that the San Andreas fault is there (if you did not have the mapped trace of it already available to you in the GIS project)? Can you see evidence for any other major fault? Where in this region do you recommend that our class do geophysical surveys across the San Andreas fault? Why do you suggest these locations? The GIS Lab is available to you all the time. It opens with key card access after hours and on weekends. For workstation use, students doing classwork have priority over those doing research. There will be two sessions on Fri. Feb. 27 in the GIS lab: one at 11 and one at 3.
See you tomorrow in the GIS lab 309 North Mudd 11-12 a.m. Thomas Vanessa Zhongwen Dongzhou Lori Sara 3-4 p.m. Nina Veronica