![]() ![]() Of the three years, 2000 was the most different, with a greater percentage of the global land scenes acquired two to seven times in 2000 compared to the other years (Fig. 2a) i.e., 97 – 98% of the scenes were acquired at least once (Fig. The LTAP goal of acquiring every land scene at least once is evident in that only 2 – 3% of the global land scenes were not acquired in a given year (Fig. The annual acquisition frequency is quite similar among the 3 years, and the annual 2002 results broadly reflect the other two years. 2 summarizes the acquisition frequency, showing the percentage of the global land scenes that were acquired annually in 2000, 20 (left column) and over each of the four seasons in 2002 (right column) a certain number of times (top row) and at least a certain number of times (bottom row). Similar distributions to those illustrated in Fig. The preferential acquisition of consecutive scenes falling along the same Landsat path is evident in certain regions, particularly those with no or a small number of 2002 acquisitions (seen as vertical stripes in Fig. The majority of the land scenes in Eurasia, Northern Canada, Northern and Southern Africa, and Australia are acquired no more than 10 times. Boreal regions and many deserts are acquired less than 5 times, and oceanic island and high latitude scenes are acquired only once or twice. Typically, Equatorial regions and some parts of central Asia, and Europe are acquired between 11 and 21 times, although certain tropical forests are acquired the maximum number of times. Northern Alaska has fewer acquisitions than the maximum because of high Winter solar zenith angles. The maximum number of acquisitions per year (22 or 23) is evident over the CONUS and Hawaii. 1 illustrates the global land scenes and the number of daytime acquisitions per scene in 2002. ![]() This may result in the over- or under-estimation of P one+ and P two+ for individual pixels that compared to the majority of the scene pixels are persistently cloudy or cloud-free. P one+ and P two+ are computed from scene-level cloud fraction metadata and so are average probabilities for all the pixels in the scene they are interpreted as the probabilities of acquiring at least one and at least two cloud-free observations of a Landsat ETM+ pixel within a scene respectively. The longest 80-day duration corresponds to the maximum effective season length when the number of ETM+ acquisitions is considered a scene maybe sensed up to 6 times in an 80-day period. In this study we consider durations of 16, 32, 48, 64 and 80 days. D is the set of permutations containing cloud-free observations occurring no more than the duration of interest apart.
0 Comments
Leave a Reply. |
AuthorWrite something about yourself. No need to be fancy, just an overview. ArchivesCategories |