X-B8, 2012 International Archives of the Photogrammetry, Remote Sensing and Spatial Information Sciences, Volume XXXIX-B8, 2012 
XXII ISPRS Congress, 25 August — 01 September 2012, Melbourne, Australia 
function files containing smooth renditions of the original data Taquari and Paraguai), because they are inundated most of the 
  
| NDVI prod (Figure 1). year. 
product 
The MOD09GQ | 240 : : : : 59°00" W 57°00°W 55°00"W 
ery including the | . té lenght » vU 
376 nm) and the | 2 M 
      
      
     
te of the highest- | 180619 
ides Normalized | 
nced Vegetation 
R), mid-infrared 
ere obtained from 
gov/) as a free 
tion Tool Web 
ts°g'0"S 
    
    
  
  
TEE) 
   
      
  
       
     
      
oss 
SR 
$$ 
2 
on 
tm. 
+ 
e 
   
Cn 
e 
ph 
» 
NE 
RR 
d 
+ 
M 
L2 
+ 
5 
25 
tt 
X 
    
  
scaled NDVI 
Ht 
ER 
Fics 
t 
TS 
Fees 
en 
  
  
  
ages were subset | Ho 
ing the WGS84 | 130} Integral ® 
processing was + "En 
AESAT. | ^7» — 4 5% 60 10 1,40: 00.10 
d is an evolution | time 
n which Pantanal | Figure 1. Seasonality parameters generated in TIMESAT 
avior, taking into (Modified from Jônsson and Eklundh, 2004) 
behavior of the | 
ightness, textures 
Irought (obtained 
Based on the granulometric analyses the samples were divided 
into four groups with different textural compositions, based on 
ted from digital | the amount of fine-grained sediments: Group 1 — samples ES 
data. The new containing more than 70% of fine-grained sediments; Group 2 — 
done in ArcGIS samples containing from 50 to 70% of fine-grained sediments: 
g the behavior of Group 3 — samples containing from 30 to 50% of fine-grained 
region. Then, we sediments and Group 4 — samples containing less than 30% of 
imum, maximum fine-grained sediments. Fine-grained sediments are described 
5 for NDVI from | here in terms of their silt and clay content. 
For each group we generated the time-series of the mean NDVI 
  
       
  
  
(MOD13Q1) was values by averaging the values of all granulometric samples sus 
10 (115 granules) | identified in one of these groups. Then, we calculated each 
or granulometric group statistics mean, standard deviation, minimum, maximum soto" w STOW — 
ship between the and range for NDVI. After all the annual seasonality parameters 7 : 
and responds to generated by TIMESAT for the samples were converted in 
trations (mostly mean values (mean of the 5-year seasonality parameters) for 
ries of this data each group samples. Hoa 
xy — of vegetation 
Is act as substrate 3. RESULTS 
t its composition | 
ion studies. The The two images obtained from MODO09GQ allowed the 
iformation about subdivision of Pantanal in 20 homologous areas (Figure 2). 
The triplets with the Red and NIR bands and vegetation index 
| for atmospheric built on these bands, despite the spectral poverty, revealed to be 
z the data quality, tonally rich, allowing us to observe the different behaviors of HOS 
sidual clouds and vegetation in both seasons. 
)7). To minimize Since NDVI tells essentially about the photosynthetically active 
e Savitsky-Golay radiation, this parameter presents great importance in the 
on and Eklundh, differentiation of sub-areas of the Pantanal to inform, 
es profiles were ultimately, about water availability in non-flooded parts of the 
int. The MODI3 region. 
| point in the time In Figure 2, it is possible to observe that some regions remain 
(1.0), values 1-2 With a similar vegetation behavior (yellowish areas in both 200s 
0.5), and value 3 | seasons), remaining with vigorous photosynthetic activity even 
fitting parameters | in the dry period, while others have low photosynthetic activity 
Iter procedure, 4- | in this period (bluish areas in the dry season). The areas that 4 Ghootes SEO Pantaneiro 
strength of 2.0, no | appear bluish during both seasons are dominated by sandy 03. Cuba = A E 
.0, and begin and | sediments and poor soils and have lower NDVI values, while 0e Eco uno 16 Ba eins 
areas that are reddish stay soaked most of the year. 06. Baräo de Melgaço 16. Alto Aquidauana 
M : * 3 07, Paiaguas 17. Planicie do Negro 
containing annual The analysis of the pair of MODIS triplets showed that the 08. Taquari 18. Nabileque 
the season; Base variations in mean NDVI values within each area are closely 99 Ana Nec CHR ir Min p22°007$ 
values; Peak time: related to the predominant granulometry, ranging around 10% t Ca { 
z the season; Peak at the argillaceous areas and between 20 to 30% at the sandy Figure 2. Limits of Pantanal on MODIS triplets of the rainy and 
iction during the areas (Table 1). Only three of the 20 areas analyzed filled with dry seasons, respectively (False colour composite: 
as well as fitted coarser sediments didn’t show these trends (Planície do Negro, NDVI/NIR/R)