Landsat and MODIS images are good choice for your application.

To detect a fire both MODIS and Landsat TM might be good choices, however, MODIS has very low spatial resolution compared to Landsat TM (250m vs 30m).  BUT MODIS has a better temporal resolution compared to Landsat TM (daily vs 16 day repeart cycle).  If you want to monitor a fire you probably need at least daily imaging, but would be better if 2 to 3 times a day, which means even MODIS is not acceptable.  Landsat TM is not acceptable for monitoring, however, with the spatial resolution makes it a very good data set to detect and analyze the damage after the fire.  The SWIR bands (TM bands 5 and 7 in Landsat 5 and 7 systems) are very useful for detecting and mapping burnt areas (and how badly burnt they are).  If you are lucky enough to have a Landsat TM image during a fire the SWIR bands are again very useful in that they penerate thru moderate to some what heavy smoke and the areas with active flames can be seen.

Pat

what about SAR data and the use of AVHRR ,ASTER images for detect fire and identify burned area?

MODIS

MODIS

MODIS

What is your major requirement? Higher temporal or spatial resolution? While MODIS gives you a better temporal resolution Landsat will be preferred if you want to trade temporal resolution over spatial. I've read from answers above suggesting landsat TM. But if you decide using landsat the new landsat 8 thermal bands may be better due to their higher pixel depth.

since you are only interested in identifying the areas affected by forest fire, Landsat is a good choice

MODIS

Dear Mohamed, 

Very relevant suggestions have been given to your question. I would agree most with what were suggested by Prof Chavez and Dr Charma. I would just give some complementary points to those of others.

In my opinion, you can use both MODIS and Landsat data for your purpose. It is really difficult to monitor real-time burning due to lack of data. Geostationary satellite data can meet the requirement of "temporal resolution" but their spatial resolution is too coarse to be useful. So what you can do most is to identify the burnt area and assess the damage. For local study, you may use Landsat data, however, if your study area is whole Asia, you have not much choice but MODIS data. In this case, you can use a multiscale approach, that is, using MODIS data for regional-scale detection and Landsat data for local-scale detection and verification. The advantage of these two satelite data lies in that both of them have visible, shortwave and thermal bands, which allow you to produce NDVI, EVI, burnt ratio index (using TM band 4 and 7) and land surface T. 

My experience shows that there is shortcoming in MODIS data due to cloud mask. So I would not suggest you to just use a single date or a single combination of MODIS data (8-days or 16-days NDVI, EVI, or other LST) for your study, but using multiple dates or multiple weeks (e.g., 4-6 weeks) to extract their maximum or minimum value of each pixel to avoid cloud problem, which is a common problem in subtropical and tropical areas when applying optical remote sensing. 

For the most cloudy area, Radar data is a good choice. Sentinel is free of charge, you may use it for your fire detection purpose at local-scale.  

Good try and best,

Weicheng     

Dear Mohomed,

as answered before, the using of satellite data doesn't depend on the region: all of them are global and have an excellent spatial reference (they're georeferenced using the net of terrestrial control point for each scene, so the problem could occurs in difficult conditions like mountains or great water areas with the lack of terrestrial control points, but not because of the region). I just would like to add some new informations, probably it also helps:

1. the Landsat better for mapping burned areas or for positioning of the burned area, but not for fire monitoring. I recommend to use Landsat 8 data (because Landsat 5 is not working after November 2011 and Landsat 7 has the stripes of No Data; Landsat 8 has the best radiometric resolution also); NIR and SWIR bands are fit you task.

2. If your research required the fire monitoring within 1-2 days or if you need to process the vast areas, the best solution is MODIS data. Accompany to rather data you can use MODIS subsets https://earthdata.nasa.gov/earth-observation-data/near-real-time/rapid-response/modis-subsets and even FIRMS data https://earthdata.nasa.gov/earth-observation-data/near-real-time/firms. The last dataset is based on MODIS data and can be downloaded in *.shp and other formats (it's handy to use it together with Landsat data or just to assess the extent and places of fires).

3. In my opinion, the tandem Landsat 8 OLI - MODIS is the best for fire monitoring and assessment (you can choose one or other depends on what resolution is more important: spatial or temporal. Aster, Spot and other availiable data have some disadvantages in temporal or spectral resolution). But I didn't check the new Sentinel 2A data yet (https://sentinel.esa.int/web/sentinel/sentinel-data-access); taking in mind it's features it looks like to be very important source of data you may need also.

MODIS

Dear, good morning to all,

I have a question, how and with what program can I open the img Sentinel 2 ???

I have dealt with Envi3, I could not; with QGis, not everything I could not; With SNAP, I do not know how to lift them.

The idea would be to open them as the OLI, invoking a spindle / metafile, and raise all the bands within each subsystem, but we know how to do it.

Grateful if you can give me a hand,

thank you,

danieljp

E-Mail: