TABLE OF CONTENTS
Earth Observatory vs Goddard: NDVI Data
If we try to locate comparable information to that in the EO site, e.g. to its Life on Earth: Vegetation data, which shows the greening of the earth with the seasons and the extent of polar snowfields in the north and south year by year, we can select Land Biosphere from the DAAC menu and are taken first to the option to browse and select the data we want by means of a search page that has drop-down boxes like those on the EO visualization page (Figure 3a, compare Fig. 1b) listing the various parameters available (in DAAC these are datasets, but classified by their typical scientific uses) and similarly, again like EO, for the time ranges. In the DAAC site, time can be specified to the day, not just the month as in EO, and geographical regions can also be isolated. But once a data set is selected, our choice is either to download it by FTP (the faster forerunner of HTTP on the web), typically 35 megabytes, or to browse it … but this is only possible if we have a unix-based workstation computer, not an ordinary PC. True scientific users would then have a comparable experience to the EO visitor. Further technical literacy is required to make use of the instructions given on how to set up the browser function on the workstation.
Figure 3a. NASA EOS Data Gateway: Data Search and Order Form Page
(Access via link: Enter as "Guest" from http://harp.gsfc.nasa.gov/~imswww/pub/imswelcome/plain.html)
Unlike the EO site, browsing images does not include brief color-key codes and an explanatory paragraph. Rather, there is an entire menu of resources and documentation to assist the technical user in interpreting the data displays. In some cases this includes links to the researchers who have created the datasets; and in all cases links to the published scientific literature describing all aspects of the production of the data. If we in fact look at some of this documentation (there is a 96-page printable manual you can download for the data we are interested in here on vegetation cover), we find after three more jumps a sample of the displays of our data, in low resolution, as a grid of maps, very similar to what we can get in the EO visualization (see Fig. 3b). It is accompanied by a reference to the published literature, and the key explanation:
The Climate Data Set is a 1 degree by 1 degree product for use in global, coarse resolution modelling studies. The Climate data set is a global, 1-degree x 1-degree, NDVI field. This is derived from the Composite Data Set by calculating a mean Channel 1 reflectance and a mean Channel 2 reflectance for each 1-degree area in which 50% or more of the 8 km pixels are identified as land pixels.
Figure 3b. NASA GSFC Global Land Biosphere Data and Resources, Climate Data Set Webpage
(January 2000, http://daac.gsfc.nasa.gov/CAMPAIGN_DOCS/LAND_BIO/Climate_ds.html)
There is no automatic glossary function here as there is in the EO site, but we can use the online documentation to unpack the meaning of much of the technical language. In doing so, we get some sense of the multimedia literacy demands of professional scientific genres. For example, ‘composite data set’ leads back to ‘daily data set’ and its description is primarily in the form of a table:
Daily Data Set Layers
Layer Units Range----------------------------------------------------------NDVI - -1 to +1CLAVR flag - 0 to 31
QC flag - 0 to 31
Scan Angle radians -1.047 to +1.047Solar Zenith Angle radians 0 - 1.396Relative Azimuth Angle radians 0 - 6.283Ch1 Reflectance % 0-100Ch2 Reflectance % 0-100Ch3 Brightness Temps Kelvin 160-340Ch4 Brightness Temps Kelvin 160-340Ch5 Brightness Temps Kelvin 160-340Date and Hour of Obs DDD.HH 1-366.23
The actual parameter in the data
set which underlies the images of interest to us, and which is in fact
identified, but not explained, even in EO is ‘NDVI’. Following on to the
basic documentation for this we find:
The Normalized Difference Vegetation Index (NDVI), which is related to the proportion of photosynthetically absorbed radiation, is calculated from atmospherically corrected reflectances from the visible and near infrared AVHRR channels as:
(CH2 - CH1) / (CH2 + CH1)
Where CH1 is the reflectance in the visible wavelengths (0.58-0.68 um) and CH2 is the reflectance in the reflective infrared wavelengths (0.725-1.1 um). The principle behind this is that Channel 1 is in a part of the spectrum where chlorophyll causes considerable absorption of incoming radiation, and the Channel 2 is in a spectral region where spongy mesophyll leaf structure leads to considerable reflectance (Tucker 1979, Jackson et al.1983, Tucker et al. 1991).
This paragraph is perfectly typical of mathematical-scientific register: the set-off algebraic formula expression, followed by the ‘Where …’ definitions, the citations to the literature, as well as the use of technical terms, and embedded numerical expressions with units of measure. Note the hypertext link on ‘reflectances’ which is a new feature of this medium. It leads in fact not to a simple text definition, but to a complex page dominated by black-and-white images showing the differences between the channel 1 and channel 2 data used in NDVI calculations; this Reflectances page also contains embedded quantitative data with measure units and a link to the published literature.
The literacy demands here are quite comparable to
those of scientific print publications (Lemke 1998a), but go beyond them in the
specific matter of hypertext (or hypermedia) literacy. In print genres, the
scientific citation is a standard intertextual referring device; to make use of
it requires not just language literacy skills to interpret the citation, but
also the activity skills needed to physically locate the printed text referred
to. That latter demand is being simplied by the hypertext links of web-based
genres (although it is still present here; the links to the citations lead not
to the original papers but only to the detailed citation information in an
online bibliography). At the same time, it becomes easier in the new medium to
increase the density of links, and so the complexity of the intertextual web
that users must integrate in order to make full sense of any part of it.
Hypertext literacy is a large subject in itself (e.g. Rouet & Levonen 1996, Reinking et al. 1998), raising basic questions of how people navigate through large hypertextual webs, and how we make meaning across both short-term, small-scale trajectories of sequentially linked webpages and also longer-term, larger-scale trajectories which may eventually include a large fraction of all the locally linked pages at a site (or within a meta-site). Hypertext literacy is feature of the emerging extended literacy of computer-based multimedia ( Lemke 1998c), which happens to have its leading developmental edge in the domain of scientific communication.