I'm going to stick my neck out and answer the question in one word. No.

Now to qualify this.

Brian's answer sets you on the right track, with phenomenology as the first step. Can the sensor you are using detect the effect you are looking for? Immediately you need ancillary data, even if it is simply descriptive of the phenomenon.

An approach which can work and may help you, in the way you pose the question is to look at unsupervised classification methods/PCA etc. Does this identify anomalies around your landfill? Now you have anomalies. Nothing more. In some cases you can characterise them using your data, but eventually you are going to need to get onto the ground for some sort of cal/val exercise.

I'm answering this as I have experienced a number of studies where remote sensing has been employed to produce results which look great, while not answering the question which needs to be answered. It's important to guard against this as it is bad for the image of remote sensing.

Hello Mam,

here i have added few poThe findings of this study are:

(a) LISS-III image products can provide accuracies comparable to Landsat,

(b) operational processing issues with using IRS data have been identified and can be resolved

with the expertise available; and

(c) hence LISS-III imagery, if available, is a feasible substitute for Landsat within the Land Cover

Change Program.

The recommendations for the Land Cover Change Program are:

• Individual ‘map’ oriented LISS-III and AWiFS imagescan be adequately registered to the AGO

base images with minimal additional ‘per image’ effort. Additional Master GCP features may be

required for some images as the existing features used with Landsat images may not provide a

sufficiently dense spatial coverage for the slightly smaller LISS-III scenes. The ortho-rectification

effort will increase by a small ‘per image’ amount corresponding to increasing the number of

Master GCPs, as well as by the increased number of images to be processed.

• Calibration of the IRS imagery requires extensive further study. New BRDF kernels may need to

be derived as well as new model coefficients. As the intensity range of the LISS-III and AWiFS

imagery is much greater than for the Landsat TM imagery, a new calibration base will need to be

investigated, particularly if the data is to be later used to monitor trends. As with the Master

GCPs, the invariant target locations used in the calibration process may be too sparsely located for

the smaller images. Once appropriate methodologieshave been established, the operational

processing effort should only increase by the ‘per image’ amount necessary for the selection of

new invariant targets.

• In the thresholding stage of the processing, new indices will have to be derived for most

stratification zones as the not all Landsat TM spectral bands are present in the IRS image data.

This may take up to two weeks of effort per map sheet. The analyses should be performed by a

very experienced team. Ideally indices are derivedby considering two or more image dates to

ensure they are robust through time rather than tailored to particular conditions in a single image.

All indices derived for the first epoch using IRS data should be reviewed when a second epoch is

available. Once indices have been established, the overall thresholding effort is increased

compared to Landsat TM data, in the worst case by the number of extra scenes required for

complete coverage.

• Over stratification zones not containing black soils, forest cover extent and change products from

the LISS-III are more consistent with those obtained from Landsat 5 or Landsat 7 SLC-off data

than are those from the SPOT 4 data tested. The rates of change between the LISS-III epoch and

the rest of the sequence appear similar to those from the Landsat 5 and Landsat 7 SLC-off

sequences. However, on black soil zones, the cover types where commission errors and ‘scattered

pixels’ are mapped generally differ.

• The forest extent and change products from the AWiFS data show significant additional areas of

change on small or narrow features and the edges ofmost larger forest regions. This is to be

expected as a consequence of its coarser spatial resolution, and the observed effects are similar to

the differences between Landsat MSS and TM epochs.For this reason, LISS-III imagery would

be the preferred option.

ints

If it is methane emissions, I think methane will affect vegetation, so vegetation mapping change mapping may be useful, although the resolution may be too coarse.

There are minimally four things to consider when using remote sensing imagery for a study of this sort, in order of decreasing importance: 1) phenomenology; 2) spectral resolution; 3) signal/noise; and 4) spatial resolution. Phenomenology dictates whether the material of interest can be detected and differentiated from other materials in your area of interest. If not, then you can’t do this no matter what. Next is spectral resolution. Does your sensor have sufficient spectral resolution (nm vs um) to take advantage of the phenomenology, i.e., detect your material of interest, either primary and/or proxy materials. You need this so some feature(s) can be classified. Signal to noise ratio is important because a noisy detector is almost useless. Look for s/n over 100; the larger, the better. Lastly, and maybe least important, do you have sufficient spatial resolution, i.e., 1m vs 1 km pixels? If you can, let the phenomenology drive sensor selection.

To follow up on my friend's (Dr. Brian Penn) answer, I would advise taking a sub-scene that covers about twice the area of the landfill and performing a principal component analysis, using all the bands of LISS III or IV. There will be n principal components when there are n bands, and the Principal Component Anlaysis will show you the areas on the ground that have the most spectrally unique signatures, particularly in the one or two highest-numbered PC's, where the areas on the ground with the most exotic spectral features exist (assuming there is not much noise in the scene from the detectors). When you find a particular color feature that looks like it could be a drainage from the landfill (from the exterior or the interior of it), bring up a natural color image of the scene and use image processing software to extract the values of each of the spectral bands for 10 or so pixels that are members of that specially colored high-numbered Principal Component feature that you have extracted the LISS spectral band Digital Numbers (DN) for. Then make spectral ratios of the single bands, like I have discussed in my text book, Fundamentals of Geological Remote Sensing, Prentice Hall Publisher, in Appendix C of that book. I have calculated the ratio codes of many minerals and rocks for LANDSAT TM in that book, for which used copies are available at Amazon.com (I get no royalty from these used book sales). I have not calculated it for LISS spectral bands, but you could either do so, or you can use only the LISS spectral bands most like LANDSAT TM bands and calculate those ratios and use the LANDSAT TM spectral ratio code tables to determine what minerals they most represent. I would particularly suggest looking for Jarosite (ferric sulfate), which comes from the oxidation of pyrite and is orange-red in color. It usually is found in acid mine run-off from abandoned coal mines, and might be in landfill plumes, also. That's the best I can do for you in this answer space. Best of luck. Dr. Robert K. Vincent (Bob), Prof. Emeritus, Dept. of Geology, Bowling Green State University, Bowling Green, OH.

I go with Mr.Graham Andrew Moore (methane emission), because by using temporal data study ( Data taken of the same area with different time i.e. old and present) of the particular landfill we can understand the impact of the landfill in that particular area. The remote sensing data shows the variation among the two images (old & present) this results shows the effects of the landfill. Like earlier the green vegetation coverage, presence of water bodies, etc. From these we can extract the ancillary information.

I'm going to stick my neck out and answer the question in one word. No.

Now to qualify this.

Brian's answer sets you on the right track, with phenomenology as the first step. Can the sensor you are using detect the effect you are looking for? Immediately you need ancillary data, even if it is simply descriptive of the phenomenon.

An approach which can work and may help you, in the way you pose the question is to look at unsupervised classification methods/PCA etc. Does this identify anomalies around your landfill? Now you have anomalies. Nothing more. In some cases you can characterise them using your data, but eventually you are going to need to get onto the ground for some sort of cal/val exercise.

I'm answering this as I have experienced a number of studies where remote sensing has been employed to produce results which look great, while not answering the question which needs to be answered. It's important to guard against this as it is bad for the image of remote sensing.

In my opinion the change detection of landuse land cover will show the different reflectence value because of pollution and also reflected in canopy cover so, one can go for it. High resolution data can be used for it.

Iam agree with mr. Suneet view that we can depend on the study of change detection of land cover & landuse

The use of multispectral satellite data for understanding soil pollution is indirect. The effect of soil pollution if it is significant can be understood using temporal data sets. If the pollution has significant impact on the landuse / landcover, a change metric of this parameter can highlight this effect of pollution. Ideally hyper spectral data sets can be good source for understanding the soil chemistry around the landfill site.

Dear colleague( kajori).

satellite imagery will assist you to determine type of soil. Because in soil pollution is very crutial to understand the soil type. Also sattellite imagery specifically to.. can add to you other pollutant present.

satellite imagery can be of important to give you the geological example water table and also flow direction of pollutant in water and will assit you to predict the impact in future at a certain location.

Note the satellite using the sensor with the following can be useful.

a. mutltitemporal scaner

b hyperscpectral or if not available use multispectral scaner band

c high resolution

all the best kajori

I think the better, it´s possible You can make a workfield in the local. Other possibility is a consult the database official the study area.

Kajori previous researchers have rightly indicated. But youy can start with your own geology and geomorphology map which can be easily generated using the LISS III/IV data. The other thing which is disagree is if you are looking at soil pollution much of the chemical pollution will not be seen in satellite dataset. So you have to rely on the laboratory determinations. Regarding Methane, much of it will be lost to atmosphere, rather than absorbed by vegetation and even if vegetation is thicker around teh siet is not indicative of methane only, it can be due to availability of water as well.

The health of the vegetation on and near the landfill site may provide a proxy indication of the presence of pollutants in the soil and groundwater. Chemical excess or deficiency changes the optical properties of the leaf surface such as colour, shape or size, caused by chlorosis or wilting. This would affect transpiration efficiency and consequently leaf temperature would increase. This is potentially detectable using far infra red or thermal infra red imaging. However, I agree with Henk that some validation by ground survey is required to check samples of the vegetation for symptoms of stress. Best of luck!

I agree that multiple approaches are necessary because both longitudinal studies of field search (direct soil samples for analysis) and satellite images in a wide range of time (20 years or more) can be complementary. Just satellite images analyses cannot provide whole understanding of pollution patterns (mainly if you have different sources of pollution), and depending on the type of landscape (whether you have more or less vegetation covering the selected area), satellite can even be useless for a fair assessment - unless you select proper images and apply specific techniques as those of PCA, explained by Dr. Henk Coetzee. Wish you best of luck.

My small NGO spent two years lobbying Brussels to adopt at least US EPA MSW landfill sealing system specifications and also appealed to the European Ombudsman (unsuccessfully, during the Santer administration) for access to a Portuguese landfill's technical documents.

So our landfill pollution concern has been confined to SUBsurface leachate contamination and Kajori Parial's request leaves me wondering if that has been adequate. More to the point, others have questioned whether satellite data will serve her aim; and high-resolution imagery is expensive. So I can only echo Dr Shepherd's suggestion that neighbouring plants be used as proxies.

In this respect, perhaps the Portuguese facility is relatively clean - there is little industry in the Algarve. Or perhaps it is well-managed. Certainly, the fill strives to reduce the smell of decomposing waste and extensive infrastructure captures and generates electricity from resultant gases. But on-site smells are strong and when asked to specify plants with which to conceal cell-covering membranes I felt qualms about subjecting anything to those conditions (although gulls and storks love them). But in fact the vegetation, planted in landfill-made compost, watered with treated leachate and liberally spattered with guano, has generally thrived. However, because of the underlying membrane roots are not subjected to methane which can kill many, but not all, plants.

Ultimately, and more onerously, multiple soil analyses may be required. If the location has prevailing winds, pollution would probably be highest to leeward.

I agree with the comments above.  Hyerspectral images are useful in detecting soil pollution.  But soil samples need to be gathered in order to determine the types and concentration of chemicals present in a given location. 

i agree with james smith...

James. I've recently concluded a project where we made extensive use of hyperspectral images to map contamination from coal mining. One of the conclusions I've reluctantly come to is that we didn't learn that much more than we had learned from using WorldView II data. More of a concern was that both required quite extensive ground-truthing and that the ground truth work, in combination with high-resolution aerial photography and LiDAR was possibly more cost effective when it came to rehabilitation planning.

The best way preparing thematic maps would be ascetain the availability of Thematic maps on 1:50000 scale that would be avaialable the State dept of Geology and mines. you can make use of the this map in preparing geology and geomorphology map. Fom this map you can subset your study site.

Step 2. Try to collect samples from the locations and analyze for the soil parametrs/ pollution levels alongwith GPS points.Such parameters can be spatially mapped in using interpolation techniques.

This reference point can be identifed from satellite imagery where you can try some of the pollution indices with varying spectral bands.

Mapping data (here soil data) using interpolation techniques requires several preliminary hypothesis and is an incremental process to develop a conceptual model that can justify parameters for a sampling plan and give some value to an interpolation. Multi-source data as remotely sensed ones might be useful to support such a model of the organization/distribution, and spatial continuity of variables in space, and define population and sampling units. Depending on dates of acquisition, imagery at 6m res from LISS III/IV could be interesting in that regard.

Dear Kajori, I support Kumaranchat Vinod Kumar's statement.

 Even though, NDVI  will be a  good tool for your experiment, an attempt can be done with multi-temporal imageries.  similar indices can also used.  The results along with ground investigations will help to deliver some idea about direction and extent of soil pollution, if the soil pollution cause any significant effect on nearby flora. Julian day, Rain fall etc will be influencing factors in this.  I suggest you to choose LISS 4 data from your cart.