wind speeds were higher. Thus, the wind speed data were
utilized in the experiment, as shown in Figure 7, to determine
whether the wind speed affected the differences between the
air temperature and the retrieved
IST
/
MOD
29 product in this
research. A statistical analysis of the correlation between wind
speed and the differences between the air temperature and the
retrieved
IST
/
MOD
29 product is summarized in Table 5.
The Pearson’s correlation coefficients give a value between
1 and −1, where 1 and −1 indicate totally positive and nega-
tive correlation, respectively, and 0 indicates no correlation.
In Table 5, the results show that the differences between the
air temperature and the retrieved
IST
/
MOD
29 product were
both positively correlated with wind speed
.
In the above experiments, the proposed method exhibited
a better
IST
retrieval performance for the Zhongshan Station
data. Similar experiments were implemented using the
MODIS
-
based results and the
AWS
observation data from the Ross Ice
Shelf. The results are shown in Figure 8
.
The differences between the air temperature and the
retrieved
IST
, and those between the air temperature and the
MOD
29 product, for each
AWS
, are presented in Figure 9.
According to Figures 8 and 9, the retrieved
IST
s and the
MOD
29 product present a better agreement than the
AWS
IST
s.
The corresponding accuracy of these experiments is summa-
rized in Table 6
.
The improvement in accuracy achieved by the retrieved
IST
results can be clearly observed in this table. The wind speed
(a)
(b)
(c)
(d)
(e)
(f)
Figure 9. The differences between the air temperature and the retrieved IST, and those between the air temperature and the MOD29
product, for the Ross Ice Shelf: (a) Carolyn AWS, (b) Elaine AWS, (c) Gill AWS, (d) Margaret AWS, (e) Schwerdtfeger AWS, and (f) Vito AWS.
PHOTOGRAMMETRIC ENGINEERING & REMOTE SENSING
November 2015
869
