One of the most important discoveries in the field of earth observation is the discovery of spectral indices. At dawn of the development of this branch, it turned out that in many cases the analysis of spectral reflectance in certain spectral zones was not enough. In some cases, the ratios or differences of reflection of solar radiation in different parts of the spectrum are more informative. The indicators that characterize the relation between reflectance in various spectral channels are spectral indices. Initially, the practical application of indices was found in remote sensing of vegetation cover. The spectral indices used to estimate the vegetation state are called vegetation indices.
The first vegetation index, RVI (Ratio Vegetation Index), was developed in 1969 by Carl F. Jordan. And in 1973, John W. Rouse et al have created the Normalized Difference Vegetation Index (NDVI). NDVI is the most well-known and most popular vegetation index. It has been used in many scientific works. The abbreviation NDVI almost became synonymous with the term “vegetation index”. Therefore, if a vegetation index is referred to without specification, most likely it is NDVI. But many vegetation indices have been developed. In addition to vegetation, spectral indices have been used in studies of water bodies, soils, snow cover, surfaces of artificial materials.
If we are going to use the spectral index method, then we face a series of questions:
- which index to choose for solving a specific task?
- what is the formula for calculating the index?
- what is a specific index used for?
- which indices can be calculated based on the data from a particular satellite?
- which channels of a particular satellite sensor are used to calculate the index?
These questions are not easy, because scientists have developed a lot of indexes. Their number is already in hundreds. It is difficult to find one’s way in this variety. And all the information on the spectral indices that exist today cannot be remembered by one person. Therefore, the need for its systematization is beyond doubt.
There is a tool that sorts information about all the spectral indices existing today. It is the Index Data Base, abbreviated to IDB.
The idea of its development belongs to a group of German researchers – Verene Henrich, Gunther Krauss, Christion Götze and Christofer Sandow. They have been working since 2009, and in 2012 the project website was launched.
The IDB logo is an allusion to the formula for calculating NDVI.
Fig. 1. IDB logo
The IDB base is quite simple in structure and easy to use. But its volume is large enough. It contains information on 519 indexes, 167 satellite sensors, 43 ways of index application, and 230 references.
The IDB’s homepage has links to search queries that help you find answers to questions we have written above. Eight request forms are provided: three for searching the indexes (by sensors, by application and by sensor and application together), three for searching satellite sensors (by indices, by application, bands of sensors), two for searching index application (by indexes and by sensors).
In addition to queries, there is another way to look up information – looking through general lists that are also located on the main page of the IDB. There are four lists: a list of indexes, a list of satellite sensors, a list of index applications, a list of references. In addition, visualization of sensor bands and visualization of index wavelengths are available.
The IDB homepage (Figure 2) has links to information from spectrum indexes and links to search queries. The IDB has two forms of information representation: lists and visualization.
It is possible to use data from the IDB database on other sites. To do this, a JSON-API can be used.
Fig. 2. IDB main page
If we need to get information about any spectral index, we can find a page with its description either in the general list or through a search query.
An example of an index page is shown in Figure 3, we see information about the SAVI index. First of all, the index page provides general information. This is a table that contains the full name of the index, abbreviated name of the index, and the calculation formula. Below is the visualization of the spectral range that is needed to calculate the index.
Next, there is a table with a list of satellite sensors which you can use to calculate the selected index. There are columns describing the sensors (sensor names, number of channels, overall spectral range), and a column with a calculation formula for each sensor with band numbers.
As for the list of satellite sensors, there is a serious comment to it. There is confusion in the database concerning the names of the satellites and the names of the sensors. In some cases, the name of the satellite is used instead of the name of the satellite. So the sensors list is IRS-1C and IRS-1D. In fact, these are the names of two Indian satellites. And the sensor carried by them is called LISS. So, if you did not find the sensor name in IDB, try to look up the name of the satellite instead.
Fig. 3. SAVI page on IDB
Next, the third table on the page contains the list of possible index applications (fig. 4). And a list of scientific references to the index is seen below. You can download bibliographic information in BibTeX, RIS and CSV formats. This is great for researchers who use reference managers such as Zotero, Mendeley, EndNote, and others.
Fig. 4. Reference info on IDB
From the index page, you can go to the sensor pages and the application pages. So, this is another way to browse the information you need, in addition to searching through lists or using search queries.
Although the IDB was first created as a directory of spectral indexes, with the same success it can also be used as a guide to satellites and their sensors. On the pages of the IDB sensors, you can find information about when the satellite was launched into orbit, the type, height and inclination of the orbits (Figure 5). The satellite operator is indicated. Spectral bands and ranges are shown. Only the flyover frequency and radiometric resolution of the sensors are missing.
Fig.5. IDB sensor page
The IDB authors are continuing to develop their resources. Plans are to improve the existing listings – adding new bibliographic references and expanding the list of applications. Also, the authors plan to publish a fundamentally new type of information for the resource. These will be scripts in the IDL language that users can download and use to calculate the indices in ENVI. The functionality of the site, too, will change. In the future, the editorial board of the IDB promises to add a discussion forum and the ability for users to add comments to the indexes.