In: Wagner W., Székely, B. (eds.): ISPRS TC VII Symposium - 100 Years ISPRS, Vienna, Austria, July 5-7, 2010, IAPRS, Yol. XXXVIII, Part 7B 
A visual interpretation of the images was performed by trained 
interpreters in two stages: (i) the images were evaluated in 
chronological order, and at the moment in which an area was 
identified as harvested, it was assigned to the pertinent thematic 
class; (ii) after the visual interpretation was performed by the 
several interpreters, all the resultant maps were revised by a 
single interpreter (the reviser) to guarantee homogeneity of the 
interpretations. 
Next, a mosaic (thematic map) was generated for the entire 
harvested sugarcane area in the state of Sao Paulo for the 
2008/09 season. A declivity map, generated from the SRTM 
images, using the methodology described by Valeriano et al. 
(2006), was utilized to identify areas for mechanical harvest 
(<12% declivity) and areas of non-mechanical harvest (>12% 
declivity). The intersection between the two maps permitted the 
evaluation of the different harvest modes by declivity. 
3. RESULTS AND DISCUSSION 
The total harvested area increased 20.9% (680 thousand ha) 
from crop year 2006/07 to 2008/09. This increase is a 
consequence of the expansion occurring in the state of Sao 
Paulo during this time period. Table 1 summarizes the total area 
harvested with and without burning, the unharvested areas for 
the three harvest years, and the harvest type for each declivity 
class. During the 2007/08 season, 5.6% of the total available 
area for harvest (220.871 ha) could not be evaluated due to 
cloud cover obscuring the images. 
In the 2006/07 season, the area harvested by burning was 1.02 
million hectares greater than the area harvested without burning. 
In the 2008/09 season this difference was 73.6 thousand 
hectares, which represents an increase of 73.1% in the harvested 
area without burning between these two seasons (Table la). 
Therefore, the percentage of the total area harvested with 
burning decreased each season, from 65.8% in 2006/07 to 
50.9% in 2008/09. Despite the fact that the overall harvested 
area without burning increased considerably, the area harvested 
with burning did not show a considerable reduction. This 
indicates that, the majority of newly cultivated areas are 
harvested without burning. The limiting factor for the 
conversion of the harvest method is that the plots must be 
prepared for the harvesting machines. This requires adequate 
planting lines, and in addition, many areas have a declivity 
greater than 12%. Also, the vegetative cycle of sugarcane is 
approximately 6 to 7 years, and farmers do not reform the plots 
until the end of this period. Therefore, the plots currently 
harvested with the burning method should be gradually 
converted to non-burning plots or eliminated for sugarcane 
production if declivity is >12%. 
The total unharvested area increased each season, reaching 
11.6% of the total available area for harvest in the 2008/09 
season (Table la). The principle reason for this fact is that the 
ethanol plants under construction presented significant delays to 
enter in operational activity. Also, unfavorable weather 
conditions during the harvest season reduced the harvesting 
capacity (Aguiar et al., 2007). 
When considering the declivity classes, in the entire state, 
approximately 97% of the available area for harvest during the 
three seasons was located at a declivity of <12% (which allows 
for mechanical harvesting) (Table lb). However, harvesting 
with burning was predominant, especially in the areas with a 
declivity >12%. In this declivity class, harvesting is performed 
manually; therefore, the straw has to be burned to easy the 
harvest. 
Table 1. Total area of sugarcane harvested (a) per declivity class 
(b and c) without and with burning and the 
unharvested area, for the seasons of 2006/07 to 
2008/09. The percentages in relation to without and 
with burning refer to the total harvested area while the 
percentage of unharvested sugarcane refers to the total 
available area for harvest at the beginning of each 
/-o season. 
Year 
Season 
Harvest type 
Without burning 
With burning 
Unharvested 
(ha) 
% 
(ha) % 
(ha) 
% 
2008/09 
1,928,561 
49.1 
2,002,215 50.9 
514,502 
11.6 
2007/08* 
1,667,502 
46.6 
1,909,235 53.4 
164,321 
4.1 
2006/07 
1,113,855 
34.2 
2,138,408 65.8 
102,208 
3.0 
(b) 
Year 
Season 
Declivity <12% 
Without burning 
With burning 
Unharvested 
(ha) 
% 
(ha) % 
(ha) 
% 
2008/09 
1,891,845 
49.7 
1,917,719 50.3 
494,307 
11.5 
2007/08* 
1,630,825 
47.0 
1,835,907 53.0 
158,960 
4.1 
2006/07 
1,089,812 
34.7 
2,055,017 65.3 
98,877 
3.0 
(C) 
Year 
Season 
Declivity >12% 
Without burning 
With burning 
Unharvested 
(ha) 
% 
(ha) % 
(ha) 
% 
2008/09 
36,715 
30.3 
84,496 69.7 
20,195 
14.3 
2007/08* 
37,132 
33.8 
72,763 66.2 
5,367 
4.2 
2006/07 
24,043 
22.4 
83,392 77.6 
3,331 
3.0 
* For the season 2007/08, a total of 220,871 ha, 5.6% of the total area 
available for harvest could not be evaluated because of cloud cover 
obscuring the images. 
On the other hand, an analysis of the three seasons indicates that 
the percentage of the area harvested without burning increased 
each season. The areas harvested without burning reached 
49.7% in the 2008/09 season in areas with a declivity <12%. 
The same was not true in sugarcane areas with a declivity 
>12%. In these areas, the harvest without burning was greater in 
the 2007/08 season than in 2008/09 and harvesting with burning 
increased in the last analyzed season (Table lc). A total of 
11,993 ha could not be evaluated in areas with a declivity >12% 
in the 2007/08 season. If we postulate that this area was 
harvested with burning, the percentages of the harvesting 
method for the 2007/08 season at a declivity of >12% would be 
modified to 69.5% with burning and 30.5% without. Therefore, 
even in this situation, the percentage of the area harvested with 
burning increased slightly in the 2008/09 season in relation to 
the 2007/08 season, and the percentage of the area harvested 
without burning decreased. 
Despite the fact that the percentage of sugarcane cultivated in 
declivities >12% is low (3%), the percentage of unharvested 
sugarcane in this class for the 2008/09 season was greater than 
in declivities <12%.