Lectures 3 Flashcards
XPGR
Cross-polarized gradient ratio
Developed by Waleed; other techniques also used
Combinging freq and polarization for amplified melt signal
Threshold allows classification of melt
ICESat
need stuff
GRACE
Gravity recovery and
can be used to measure ice balance
Mascon==
like a pixel, done by Luthcke
Assessing ice sheet contributions to sea level rise
Mass Balance = precipt - surface alblation - discharge
- Flux Method= assess net different between mass input and mass output from observations and models of each component
Mass Balance = precept (ice cores and models) - surface ablation (models) - discharge (InSAR and others)
Snow measurement (RS)
MODIS monthly snow over with fractional snow…….
Spectral albedo
differs for different surfaces–can be used to ID
NDSI
Normalized Difference Snow Index, uses bands from Landsat ™ and MODIS. Measuring in dark forests is hard, particularly in the tropics. Sensitive to small changes in NSDI or NDVI,,,,
NDVI
define
SWE
Snow water equivalent–mount of water that comes from an icepack. Use 19 and 37 GHz. When no snow, similar. When snow, brightness temp is different.
Important for fire modeling, flood forecasting, hydroelectric power, watershed management.
Classification techniques
GENERAL CONCEPTS:
- Each pixel has DN for each band
- N-D scatterpolots of DN by band form clusters (plot band 4 and band 5 on x, y). Closest in cluster has similar spectral signature.
- Each cluster has a mean distance from origin and angle. Together, these define the class vector.
- Other stats describe shape of cloud.
PDF–define
SUPERVISED (need knowledge of what’s there):
Used to cluster pixels in an image into particular ground cover types. Many supervised algorithms are available (Paralellepiped, minimum distance, etc)
1. user picks types / groups
2. select pixels from each class (called training class)
3. use training classes to estimate parameters of algorithm
4. Classify every pixel into type
5. Produce image / map
UNSUPERVISED (don’t have knowledge of what’s there):
Used to cluster pixels in data set based on stats only, without training classes. Two algorithms: Isodata and k-means.
Fast. Does not assume normally distributed PDF.
HYDROSPHERE LECTURE
HYDROSPHERE LECTURE
Hydrosphere applications (IMPORTANT)
list from slide–add instruments
- sea / lake surface temp–thermal, PM
- wind over lakes / ocean–AM, scatterometry
- chlorophyll–vis, NIR
- sediment–vis, NIR
- ocean circulation–TOPEX / POSEDISON, Jason, Metop, AM
- volume & mass change: Altimimetry (ICESat and aircraft, Gravity (GRACE)
- water extent–vis, NIR, AM
Cryosphere applications (IMPORTANT)
list from slide–add instruments
Water Cycle
see slide
Hydrological cycle
Global water cycle
three major pathways: precipitation, evaporation/transpiration and vapor transport