Search

A multistage sampling technique, with probability proportional to size, for forest volume inventory using remote sensing data is developed and evaluated. The study area is located in the Southeastern Brazil. The LANDSAT 4 digital data of the study area are used in the first stage for automatic classification of reforested areas. Four classes of pine and eucalypt with different tree volumes are classified utilizing a maximum likelihood classification algorithm. Color infrared aerial photographs are utilized in the second stage of sampling. In the third state (ground level) the time volume of each class is determined. The total time volume of each class is expanded through a statistical procedure taking into account all the three stages of sampling. This procedure results in an accurate time volume estimate with a smaller number of aerial photographs and reduced time in field work.

A multistage sampling technique, with probability proportional to size, is developed for a forest volume inventory using remote sensing data. The LANDSAT data, Panchromatic aerial photographs, and field data are collected. Based on age and homogeneity, pine and eucalyptus classes are identified. Selection of tertiary sampling units is made through aerial photographs to minimize field work. The sampling errors for eucalyptus and pine ranged from 8.34 to 21.89 percent and from 7.18 to 8.60 percent, respectively.

The two phase sampling technique was applied to estimate the area cultivated with sugar cane in an approximately 984 sq km pilot region of Campos. Correlation between existing aerial photography and LANDSAT data was used. The two phase sampling technique corresponded to 99.6% of the results obtained by aerial photography, taken as ground truth. This estimate has a standard deviation of 225 ha, which constitutes a coefficient of variation of 0.6%.

A two phase sampling method and the optimal sampling segment dimensions for the estimation of sugar cane cultivated area were developed. This technique employs visual interpretations of LANDSAT images and panchromatic aerial photographs considered as the ground truth. The estimates, as a mean value of 100 simulated samples, represent 99.3% of the true value with a CV of approximately 1%; the relative efficiency of the two phase design was 157% when compared with a one phase aerial photographs sample.

Automatic Image-100 analysis of LANDSAT data was performed using the MAXVER classification algorithm. In the pilot area, four vegetation units were mapped automatically in addition to the areas occupied for agricultural activities. The Image-100 classified results together with a soil map and information from RADAR images, permitted the establishment of the final legend with six classes: semi-deciduous tropical forest; low land evergreen tropical forest; secondary vegetation; tropical forest of humid areas, predominant pastureland and flood plains. Two water types were identified based on their sediments indicating different geological and geomorphological aspects.

The utilization of remotely sensed orbital data for forestry inventory. The study area (approximately 491,100 ha) encompasses the municipalities of Ribeirao Preto, Altinopolis, Cravinhos, Serra Azul, Luis Antonio, Sao Simao, Sant Rita do Passa Quatro and Santa Rosa do Viterbo (Sao Paulo State). Materials used were LANDSAT data from channels 5 and 7 (scale 1:250,000) and CCT's. Visual interpretation of the imagery showed that for 1977 a total of 37,766.00 ha and for 1979 38,003.75 ha were reforested with Pinus and Eucalyptus within the area under study. The results obtained show that LANDSAT data can be used efficiently in forestry inventory studies.

Images at a scale of 1:250.000 were visually interpreted for identification and area estimates of sugar cane plantations in Sao Paulo. The basic criteria for crop identification were the spectral characteristics of channels 5 and 7 and their temporal variations observed from different LANDSAT passes. Using this technique, it was possible to map the sugar cane areas as well as the sugar cane already harvested. An area of 801,950 hectares was estimated within the study area. The confidence interval of correct classification ranged from 87.11% to 94.71%.

LANDSAT imagery at the scale of 1:250.000 and obtained from bands 5 and 7 as well as computer compatible tapes were used to evaluate the effectiveness of remotely sensed orbital data in inventorying forests in a 462,100 area of Brazil emcompassing the cities of Ribeirao, Altinopolis Cravinhos, Serra Azul, Luis Antonio, Sao Simao, Santa Rita do Passa Quatro, and Santa Rosa do Viterbo. Visual interpretation of LANDSAT imagery shows that 37,766 hectares (1977) and 38,003.75 hectares (1979) were reforested areas of pine and eucalyptus species. An increment of 237.5 hectares was found during this two-year time lapse.

Automatic classification of LANDSAT data supported by aerial photography for identification and estimation of wheat growing areas was evaluated. Data covering three regions in the State of Rio Grande do Sul, Brazil were analyzed. The average correct classification of IMAGE-100 data was 51.02% and 63.30%, respectively, for the periods of July and of September/October, 1979.

The state of Mato Grosso do Sul was selected as the study area to define the recognizable classes of Eucalyptus spp. and Pinus spp. by visual and automatic analyses. For visual analysis, a preliminary interpretation key and a legend of 6 groups were derived. Based on these six groups, three final classes were defined for analysis: (1) area prepared for reforestation; (2) area reforested with Eucalyptus spp.; and (3) area reforested with Pinus spp. For automatic interpretation the area along the highway from Ribas do Rio Pardo to Agua Clara was classified into the following classes: eucalytus, bare soil, plowed soil, pine and "cerrado". The results of visual analysis show that 67% of the reforested farms have relative differences in area estimate below 5%, 22%, between 5% and 10%; and 11% between 10% and 20%. The reforested eucalyptus area is 17 times greater than the area of reforested pine. Automatic classification of eucalyptus ranged from 73.03% to 92.30% in the training areas.

Imagery obtained with channels 5 and 7 was visually interpreted in an effort to determine the spatial, spectral, and temporal characteristics of a 105,000 hectare area of Fazenda Mutum which was reforested with various species of pine and eucalyptus. It was possible to map a reforested area as small as 6 hectare in its initial implantation using contrast with the surrounding targets. Five classes were mapped: nondeforested areas, partially deforested areas, deforested areas, partially reforested areas, and fully reforested areas. In 1979, 12,000 hectare were deforested, 4,330.83 hectare were partially reforested, and 42,744.71 hectare were reforested.

The feasibility of using automatically processed multispectral data obtained from LANDSAT to identify wheat and estimate the areas planted with this grain was investigated. Three 20 km by 40 km segments in a wheat growing region of Rio Grande do Sul were aerially photographed using type 2443 Aerochrome film. Three maps corresponding to each segment were obtained from the analysis of the photographs which identified wheat, barley, fallow land, prepared soil, forests, and reforested land. Using basic information about the fields and maps made from the photographed areas, an automatic classification of wheat was made using MSS data from two different periods: July to September and July to October 1979. Results show that orbital data is not only useful in characterizing the growth of wheat, but also provides information of the intensity and extent of adverse climate which affects cultivation. The temporal and spatial characteristics of LANDSAR data are also demonstrated.

The utilization of satellite remote sensing in forestry is reviewed with emphasis on studies performed for the Brazilian Amazon Region. Timber identification, deforestation, and pasture degradation after deforestation are discussed.

Maps indicating areas of sugar cane cultivation in the state of Sao Paulo, Brazil are presented.

There are no author-identified significant results in this report.

The author has identified the following significant results. The class of reforestation (Pinus, Eucalyptus, Araucaria) was defined using iterative image analysis (1-100) and LANDSAT MSS data. Estimates of class area by 1-100 were compared with data supplied by the forestry institute in Sao Paulo. LANDSAT channels 4 and 5 served to differentiate the Pinus, Eucalyptus, and Araucaria from the other trees. Channels 6 and 7 gave best results for differentiating between the classes. A good representative spectral response was obtained for Auraucaria on these two channels. The small relative differences obtained were +4.24% for Araucaria, -7.51% for Pinus, and -32.07% for Eucalyptus.

The author has identified the following significant results. Visual and automatic interpretation of LANDSAT imagery was used to classify the general Pinus and Eucalyptus according to their age and species. A methodology was derived, based on training areas, to define the legend and spectral characteristics of the analyzed classes. Imager analysis of the training areas show that Pinus taeda is separable from the other Pinus species based on JM distance measurement. No difference of JM measurements was observed among Eucalyptus species. Two classes of Eucalyptus were separated according to their ages: those under and those over two years of age. Channel 6 and 7 were suitable for the discrimination of the reforested classes. Channel 5 was efficient to separated reforested areas from nonforested targets in the region. The automatic analysis shows the highest classification precision was obtained for Eucalyptus over two years of age (95.12 percent).

The author has identified the following significant results. Single date LANDSAT CCTs were processed, by Image-100 to classify Pinus and Eucalyptus species and their age groups. The study area Mogi-Guagu was located in the humid subtropical climate zone of Sao Paulo. The study was divided into ten preliminary classes and featured selection algorithms were used to calculate Bhattacharyya distance between all possible pairs of these classes in the four available channels. Classes having B-distance values less than 1.30 were grouped in four classes: (1) class PE - P. elliottii, (2) class P0 - Pinus species other than P. elliotii, (3) class EY - Eucalyptus spp. under two years, and (4) class E0 - Eucalyptus spp. more than two years old. The percentages of correct classification ranged from 70.9% to 94.12%. Comparisons of acreage estimated from the Image-100 with ground truth data showed agreement. The Image-100 percent recognition values for the above four classes were 91.62%, 87.80%, 89.89%, and 103.30%, respectively.