Figure 8 Ceara Soil Associations Map 
GENERAL RECOMMENDATION 
1) The LANDSAT-TM and SPOT imagery are found more 
reliable for land use/land cover and major soil associations 
mapping and could be used effectively for any part of semi-arid 
regions of Brazil in order to submit the natural, ecological and 
environmental resources information for the development and 
management of the study area. 
2) The orbital images proved to be an extremely useful source of 
data for the purpose of detailed regional, local and rural planning 
and management and development of our natural resources. 
3) The supervised classification of SPOT and LANDSAT-TM 
data for Land Use/Land Cover mapping and accuracy assessment 
provided satisfactory results. In terms of operational reliability, the 
per pixel maximum likelihood classification of SPOT and 
LANDSAT-TM images offered the most satisfactory results in 
comparison to other classification systems. 
4) Accuracy assessment of the digital classification showed that 
some categories such as water, forest and alluvial land were 
identified more accurately than other categories. The more 
accurately identified categories may be used as framework for the 
addition of residual classes through a more conventional 
approach, such as aerial photo-interpretation. The comparison of 
digital interpretation with reference information indicated that 
digital interpretation, closely resembled to 
field observation and the overall classification accuracy was 
observed always more than 85% except the urban and rock 
outcrops mapping units.. 
5) Maps of soil associations prepared using the RECODE 
program and other information may be used for detailed planning 
and development and management of semi-arid regions of 
northeastern Brazil. 
Kennard, W.C., Teotia, H.S. and Civco, D.L. (1988). The Role of 
an automated GIS in the development and management of 
renewable natural resources of northeastern Brazil. In XVI ISPRS 
Congress, Kyoto, Japan International Archives of Photogrammetry 
and Remote Sensing, 27 (89), pp. 220-231 
LaBash, C.L. D.L.Civco and W.C. Kennard, 1989. the use of 
linearly transformed Landsat Thematic Mapper Data in land use 
and cover classification. Technical Papes, 1989. ASPRS/ACMS 
Annual Convention, Baltimore, Marryland (USA), pp. 2:53-66 
Ribeiro, G.do N. (2006), Mapeamento do Uso da Terra e 
Cobertura Vegetal em Área do Agreste Paraibano, utilizando 
Técnicas de Sensoriamento Remoto e Geoprocessamento, Areia: 
UFPB, pp 93 ( Disserta9ao de Mestrado). 
Silva Junior, E.B. (2000), Avaliafao muititemporal da atividade 
antrópica na regiao de Santa Luzia-PB com o uso de 
Geoprocessament, Areia: UFPB, pp 66 (M.S.thesis) 
Teotia, H.S. and Ulbricht, K.A. (1992), The Integration of Remote 
Sensing and GIS Technologies for Land Development and 
Irrigation Potential in the State of Cerara, NE Brazil. In: ISPRS- 
General Convention, Washington, D.C. August 2-14 Inti. Arch. 
Photogrammetry and Remote Sensing xXIX, Comm. VII, pp. 466- 
472 
585