4. Remote Sensing in the practice: Environmental applications.
NOTE. Please notice that the following links provide information on some applications only. This is because my work has been principally focused on the analysis of information retrieved by optical sensors. The reason: the free access for non-commercial purposes.
There are several application intended for remotely sensed imagery. The outcomes for some fields are promising, but for others the potential has not yet been exploited principally due to the limitations in the spatio-temporal resolution at which the data is nowadays available. The limited resources at non-profit and academic institutions is also a huge limitation; nevertheless, researches have always found a way to overcome the obstacles.
- NASA satellites global maps. I am sure that this link will show you what you can do using remote sensing techniques.
4.1 Optical Imagery.
Perhaps the most common use given to remotely sensed data has been in applications aimed to recognize the features of the landscape. Excluding the traditional photogrammetrical analysis or aerophotos, a major advancement in the analysis of images acquired by satellite platforms is the use of the capabilities provided by the multispectral data retrieved.
- An introductory Landsat tutorial. Try this link for an overview of applications where Landsat has shown to be useful.
- Landsat handbook. Details on the instrument and all the Landsat products.
- Landsat compositor, by NASA. To understand the applications of multispectral imagery in the investigation of land surface features.
- Band combinations in Landsat. Interesting summary on the use and characteristics of several band combinations.
Supervised and unsupervised classification.
After the imagery has been calibrated and corrected, the next step is to classify the data contained. There are two types of classifications: supervised and unsupervised. For supervised classification, the pixels are classified according to spectral signatures acquired in field, or from "libraries" of signatures acquired by others. On the other side, for unsupervised classification attempts, the user may simply classify the image on categories commonly equally distributed (e.g., all "red" pixels belong to category "one", "yellows" belong to category "two", and so on).
- Unsupervised classification of Spectrally Enhanced Landsat TM Data, Frankovich, 1999.
Soil and Vegetation Remote Sensing.
The soil reflectance is actually the basis under which the multispectral vegetation indices have been proposed (Liang, 2004, page 249). Vegetation is widely investigated investigated through optical imagery. Next are some links that explain the theory and application of those indices.
- The Amazon's seasonal secrets, by Dr. Fu. This article may make up you interest in the investigation of vegetation seasonal trends using remotely sensed imagery.
- Soil and vegetation optical properties, by Dr. Leblon. A short and clear background on vegetation indices.
- NDVI description, history, applications, from the Biology Department at Duke University.
- Vegetation indices Landsat in ERDAS, from the Arizona Remote Sensing Center, University of Arizona (year 2004). Tutorials on the calculation of several vegetation indices using ERDAS (Tasseled Cap transformation, Multitemporal Principal Component Analysis PCA, Normalized Difference Moisture Index NDMI, Normalized Difference Vegetation Index NDVI, Simultaneous Vegetation Index Computation Model). Try to review the notes, and the files that are attached. After this you will learn how to construct your own models using the ERDAS Model Maker.
- Landsat vegetation indices, from the Arizona Remote Sensing Centre. Details for the calculation of orthogonal transformations (Tasseled cap, PCA analysis), and ratio based indices (NDMI, NDVI, EVI, SAVI, MSAVI, SATVI).
- Enhanced Vegetation Index EVI, Terrestrial Biophysics and Remote Sensing Lab, University of Arizona.
- FAQ on vegetation remote sensing, at Yale University, by Terril Ray. Interesting.
- USGS' Image processing methods in coastal wetlands, by Ellen Raabe, R. Stumpf, OPEN-FILE REPORT 97-287.
- AVHRR Global Vegetation Index GVI. Another index.
Other applications.
- Measuring landform heights.
4.2 Multispectral imagery, thermal remote sensing, and the investigation of the earth surface radiation budget.
From my perspective, this is one of the most relevant uses given to multispectral imagery, e.g., Landsat imagery. A relevant practical application is been carried in the investigation of the surface evapotranspiration, where Prof. Allen from the University of Idaho has become an expert in the field. His site provides several links on the topic, and is worth reviewing:
- Applied remote sensing, University of Idaho, Prof. Allen. Very interesting research on evapotranspiration, wetlands.
- SEBAL North America, Dr. W.G.M. Bastiaanssen. After following his work, you will understand how useful (and profitable) can be the development of remote sensing techniques.
c) Satellite meteorology.
- Manual of Synoptic Satellite Meteorology.
d) Remote Sensing of Precipitation.
As part of the Satellite Meteorology, the investigation of the precipitation is an important application of this technology.
- Tropical Rainfall Measuring Mission TRMM. A summary on the work of Bodo Bookhagen, where have been investigated the orographic processes in the Andes and the Himalaya.
- Estimates of rain rates from passive microwave observation. A study over the Central Andes using observations from the Special Sensor Microwave Imager SSM/I.
Thus, there are several resources, and to review them all is overwhelming. As usual, I have started with the simplest, going through the next ones according to my specific needs. Then, I have entered into the contents drawn in the books, and finally into peer-reviewed papers such as:
- Papers on remote sensing (Natural Resources Canada).
Now the question is, after having found an overwhelming amount of contents: Is it necessary to continue writing tutorials? Well, as long as their material will reinforce others contributions, the answer would be yes.
No comments:
Post a Comment