The Moderate Resolution Imaging Spectrometer (MODIS), on-board both Aqua and Terra platforms, is among the best instruments at the geoscientist’s disposal. It not only has the possibly best calibration available in civilian platforms and a revisit time of 1 or 2 days, but the data is also accessible free of charge. In addition, since the Terra and Aqua platforms are designed to fly over any region at mid-morning and early afternoon, respectively, it is possible to get images from the same portion of Earth at two times of the same day. MODIS has 36 bands of varying widths spanning from visible to the thermal infrared parts of the spectrum, with ground sample sizes between 250 and 1000 m. The instrument is specially suited for regional to global applications, or for very large systems.
There are several excellent MODIS products, generated using state-of-the-art algorithms written by expert scientists, which are suited for many different purposes. For many users, these products will be just fine. However, if what we need is being able to pick up individual images, at daily frequency, and not compilation over several days (as these products often are), we need to work with Level 1 images. This is the first of a handful of posts in which I describe how to work with MODIS Level 1B images from scratch. I hope it is as useful for others as it is for me!
In this post we focus on getting the data. L1B data can be accessed from the Level 1 and Atmosphere Archive and Distribution System (LAADS).
– We open a browser and go to: http://ladsweb.nascom.nasa.gov/
– We then click on the “Data” link and then “Search”, we are directed to the data search menu:
There, we select Terra or Aqua MODIS or both, and which data set (1, 0.5 or 0.25 km) we need. In this example we select the 1km dataset. We also select the MOD03 Geolocation data, which is important for geolocating, reprojecting, clipping and calibrating the data.
– We then select the time window in which we want your data and the collection, for example the time window of importance for vegetation in the northern Hemisphere:
– We then select from where in the world we need the data (Scandinavia in this example), and whether we want day or night (or both) images (night images only available in the thermal infrared part of the spectrum). We can specify window coordinates or just draw a rectangle in the map.
– We then start the search, having also the option of saving it for future similar searches or loading a previously saved one:
– After a (possibly long) while we are presented with the search results:
The results are separated in pages, and all are marked as selected. This is very possibly more that what we want and need, as many of these images might be cloudy, so it is best to un-check all and make a narrower selection. We then press “View all”, after which we get a long list. We scroll down to the bottom and press “clear all check boxes”:
– Now begins the a possibly long process. We click in the “View RGB” for every potential candidate, checking for clouds or other undesirable things like if the region of interest is located too close to the margin of the image (off-nadir). A possibly acceptable candidate looks like this:
– We then close the preview and check the images for download:
– When we are done, we submit the order:
– We are then directed to a page where we fill our e-mail and select the data access method (if you do not know what this is, just leave the standard FTP-pull):
– We then get an order number (and an e-mail saying that) and wait for the data to be generated:
– After a while (that might be long if we selected many images that need to be generated from raw data) we get an email with the FTP address. Just connect with your favorite FTP program (I used FileZilla for this example), access is anonymous:
After patiently awaiting for the download, we now have two files in HDF format: one for the image dat set and the other (smaller) for the georeference data. The HDF image can be opened with several tools such as OpenEV, or any HDF viewer. The file is nevertheless huge and it has a very particular projection, so it is not very well suited for working in a GIS, for example. In the next post we will see how to extract subsets, reproject and convert these images to other formats.
NOTE: originally published in my old blog Tales of Ice and Stone on 2013-10-24