CARACTERIZACIÓN DE LAS CONDICIONES DEL DRENAJE Y DEL SUELO BASADA EN LA GEOMORFOLOGÍA Y EL TRATAMIENTO DIGITAL DE IMÁGENES (PARTIDO DE PUÁN, BUENOS AIRES. ARGENTINA)

Resumo

A crescente demanda de produtos agrícolas e pecuários da última década gerou, no âmbito rural, a
necessidade de melhorar a gestão dos recursos da água e do solo. Essa situação motivou um estudo
destinado a caracterizar áreas geomorfológicas homogêneas, mediante a análise da cartografia topográfica,
imagens multiespectrais e controles do terreno. A superfície corresponde ao Partido de Puán, Sudoeste de
Buenos Aires, Argentina, com uma extensão de 6.389 km² e uma população próxima a 18.700 habitantes.
Trabalhou-se sobre a base de seis grandes unidades geomorfológicas, previamente diferenciadas por
condições de relevo, drenagem e materiais superficiais, e aplicou-se o tratamento digital de imagens com o
fim de determinar novos parâmetros descritivos na caracterização das mesmas. Especificamente, a aplicação
do Color Look Up Table (CLUT) permitiu diferenciar, em linhas gerais, solos profundos de solos superficiais.
A Análise dos Componentes Principais (ACP) fez destacáveis condutos de drenagens funcionais ou
não, assim como também possibilitou agrupar corpos lacustres em duas categorias por profundidade. A
interpretação direta no Infravermelho Refletido (IR) permitiu individualizar todos os setores cobertos por água
e quantificá-los na área por unidade e para o total do Partido. Como comentário final, destaca-se a eficiente
complementação desses modos de obtenção de dados, a análise geomorfológica e o tratamento digital de
imagens multiespectrais sobre a base das unidades diferenciadas pelo primeiro. Isto torna possível estabelecer
chaves de interpretação regional, que permitam reconhecimentos rápidos para a resolução de problemas
concretos, que possam estender-se também a uma análise multitemporal.

CHARACTERIZATION OF DRAINAGE AND SOIL CONDITIONS BASED ON GEOMORPHOLOGY AND DIGITAL IMAGE DATA PROCESSING (PUÁN DISTICT, BUENOS AIRES PROVINCE. ARGENTINA)

Extended Abstract

The Puán District is located in the southwestern part of Buenos Aires province in Argentina. It is an area
of 6385 km 2 with about 18700 in habitants. Puán (the county town), Darregueira, Bordenave, Villa Iris and 17 de
Agosto are the main towns in the area. Growing concern over rural productivity during the last decade, requires
a proper soil and water resources management. For this reason, the aim of this work is to acquire new
descriptive parameters in order to characterize geomorphic homogeneous areas. For this purpose, geomorphic
units were treated together with the satellital imagery visual analysis and the digital image data processing.
The Geographic Information System IDRISI 32 was used. The geomorphic units had been previously defined
from photocartography and the ground controls. Thus, Landsat Thematic Mapper Imagery (TM), path 227, row
086 and 087, with radiometric and geometric distortions rectified, were used. The image geometry was
corrected by identifying set of features on the ground that also were identified on the image. The ground control
points were obtained by using Global Positioning System (GPS). Topographic Maps 1:100.000 and 1:50.000
scale and Satellital Charts scale 1:100.000 y 1:250.000 were used, which are included within IGM Santa Rosa
(3763) and Bahia Blanca (3963) topographic maps, 1:500000 scale. The six large geomorphic units considered
are included in two main landscapes (pampeano landscape at north and patagonico landscape at south).
Llanura pedemontana (VLlp): It is defined from isoline of 250 metres above sea level (masl), good soil/
drainage conditions, mean freatic level 10 m. Llanura noroccidental (Vllno): its boundary is defined by 250 and
170 masl altitudes, decreasing north-westward. High frecuency of ponds; soils of medium thickness and the
freatic level depht varies from 10 to 3 m. The chemical quality of groundwater is low. Llanura occidental (Vllo):
its boundary is defined by the 250 and 150 masl altitudes, sloping down westward; drainage unconnected and
highest frecuency of ponds. The chemical quality of groundwater decrease westward, the freatic level depth
varies from 13 to 6 and the chemical quality of groundwater decrease toward lowest altitudes. Soils with low
agronomic-capability because of the coarse texture and deficient rainfall input are present. Llanura suroccidental
(Vllso): it is defined below 150 masl, the topography is irregular, with important fluvial dissection; thin soils and
the freatic level depth highly variable (50 to 3 m). Paleovalle complejo (CpV): the bottom of the paleovalley
display the lowest altitudes in the area (+20 to -20 m with respect to sea level), sandy soils, scarcely developed
and freatic level depht highly variable (50 to 30 m). Niveles escalonados 1 y 2 (NE 1 y 2): these are restricted
geomophological surfaces associated with altitudes between 120 and 40 masl. Runoff not evident, freatic
level depth between 40 and 30 m. The chemical quality of groundwater is low. The first four units correspond
to the pampeano landscape and the two last ones to the patagonico landscape. The digital image data
processing produced the following results:By using the Near Infrared band (NI), the location and delimitation
through the pixel’s frequency histogram (figure 1), of the water bodies were precisely carried out (figure 2). In
the same way, the areas used by the ponds in relation with each geomorphic units and with the whole District
were estimated (table 1). The usage of Color Look Up Table (CLUT), (figure 3 and 4) allowed to differenciate
shallow from deep soils. In the same way, subdivisions in geomorphic units, difficult to differenciate with a
simple band, were identified. Bands 2, 3 and 4 constitute the best arrangement. Blue filter (spectral green), red
filter (spectral red) and green filter (Near infrared) with contrast enhacement in image data by histogram
equalization, were asigned to the bands respectively. The Principal Components Transformation (PC) of the
image data made evident features not discernable in the original data, such as the drainage way functional
character (figure 5) and changes of depth and/or sedimentary or saline load of the ponds (figure 6). A classification
starting from the image’s histogram of figure 6 (figure 7) was done. The dark tone indicates deeper water
bodies and the grey tone shallower water bodies and lands without vegetation or with low photosinthetic
activity (figure 8). The particularly components were obtained from the covariance matrix and the third principal
component (3PC) was particulary used (table 2). Finally, the application of the geomorphic main interpretation
together with multispectral images analysis, in digital and visual form, produced useful results. Thus, the
arrangement of these two groups of data provided new keys for the regional interpretation, which might be
extended to multitemporal analysis.