3

How would I know if any of the satellite imagery (Landsat, ASTER, Quickbird and so on) have DN or Digital Numbers such as integer or uninteger on its pixel. I download from the USGS EarthExplorer and have seen some in 0 -255 or into floating numbers. My question do they at USGS re-calibrated or left untouched them before they put in into their website for the user to download.

Does this means that each different imagery has its own different digital numbers from the sensor ?

I have ArcGIS 10.1, ENVI 5.0 and ERDAS 2013

closed as unclear what you're asking by matt wilkie May 2 '14 at 20:17

Please clarify your specific problem or add additional details to highlight exactly what you need. As it's currently written, it’s hard to tell exactly what you're asking. See the How to Ask page for help clarifying this question. If this question can be reworded to fit the rules in the help center, please edit the question.

  • 2
    BY DN do you mean Digital Number or something else? Please edit your question, andlet us know exactly what you are looking for. It would also help to know exactly which sensor/dataset/image you are talking about – Devdatta Tengshe May 1 '14 at 17:22
  • I am using ASTER and when I identify on the pixels it shows numbers in 0 to 255. – user29789 May 1 '14 at 17:53
  • 2
    The Q&A format of GIS SE works best with one question per Question so I recommend that you edit yours so that it is clearly focussed on the one of most importance to you. You can then research/ask the others separately. – PolyGeo May 1 '14 at 20:12
  • @user29789, what is DN? – Simbamangu May 2 '14 at 5:16
  • Have you looked at the general metadata for the particular satellite? The USGS data is very well documented. – Erica May 2 '14 at 14:55
4

You need to provide more details! There are many "satellite's". If the data is 0-255 it is 8-bit and represents DN. In a processing workflow some analyst prefer to scale floating-point to 16-bit so, bit depth does not always reflect correction level. Since you downloaded the data from Earth Explorer and it is 8-bit it is certainly DN.

I do not know what you mean by "radiation". The common correction is top-of-atmosphere-radiance and then at-sensor-reflectance. The necessity of this correction is really dependent on what you are planning on doing and how clean the image is. Any time-series comparison requires correction to at sensor reflectance, preferably with an atmospheric correction. There are other caveats as well. For example if you want to calculate a tassled cap transformation on ETM+7 the coefficients are specific to reflectance values but TM5 is calculated on DN's. If you are performing a classification or calculating indices on a single date image then it may not be necessary to perform this correction.

As far as the "expression" to apply to the data. It is dependent on the sensor. In ERDAS you need to write your own model for the correction. In ENVI there are many sensor models where you can read the header directly and apply a correction on the fly. You will need to read the help documents which, provide detailed instructions. If I am preforming time-series analysis with a mix of Landsat 5 and ETM+7 I will calibrate the dynamic range of TM5 to ETM+7 during the reflectance correction (which is not available in the ENVI TM5 sensor model).

The equations for Landsat at sensor reflectance corrections along with required coefficients are available in the USGS Landsat Science Data Users Handbooks and various other USGS websites.

Landsat ETM+7 Science Data Users Handbook

Revised Landsat-5 TM Radiometric Calibration Procedures and Postcalibration Dynamic Ranges

Landsat 8 at sensor reflectance corrections

You can easily research this on your own with a simple Google search. If you are using ASTER it is a bit more complicated and you will need to dig into the literature for the equation and associated coefficients. If you get stuck, contact me and I can provide ERDAS graphic models for ETM+7, TM5 calibration and ASTER. Although, it would be nice to see a good faith effort on your end. This site is intended to help you through problems and not just do your work for you. Based on the scope of your question it does not look like you have put much effort into researching basics.

  • This project will be a pilot crop health/ soil moisture state from a center we have and wants to do 10 years using An ASTER. – user29789 May 1 '14 at 18:10
  • In that case you will need to correct your data to radiance or reflectance. If available I would use level 1B products. The 1B processing converts 1A data to radiance and can be used for spectral analysis without any preprocessing. – Jeffrey Evans May 1 '14 at 18:42
  • When I download ASTER, they were in 1B. – user29789 May 1 '14 at 18:48
  • My bad, the radiance values are scaled back to 8-bit DN's in the 1B product, making use of a single link band coefficient (scaling factor). This allows for the equation: [radiance=(DN(i)–1)*B(i)] where; DN(i)=digital numbers for band i, B(i)=coefficient for band i. The band coefficients are in the ASTER Users guide (pp. 25-26). asterweb.jpl.nasa.gov/content/03_data/04_Documents/… – Jeffrey Evans May 1 '14 at 20:59
0

More information about what you are specifically asking is needed. However, DN typical means digital number. A DN of a satellite image is the value of each pixel of each band (channel). For Landsat images the DN will range from 0-255.

Not the answer you're looking for? Browse other questions tagged or ask your own question.