Seasonal variation in classification accuracy of forest-cover types examined by a single band or band combinations

In vegetation remote sensing, classification accuracy cannot be fixed, due to seasonal variations in spectral reflectance characteristics. This study aims to clarify the seasonal variability of classification accuracy by forest-cover type. In particular, this paper describes seasonal variability by each band or band combinations. The study area is located in the vicinities of Hisayama and Sasaguri in Fukuoka Prefecture, Japan. Natural broadleaved, coniferous plantation, and bamboo forests were studied. Supervised classification was applied to six SPOT/HRV images taken in 1997. Kappa analysis was applied to assess the classification accuracy and compare any two error matrices. The results revealed that some single band or two-band combinations were as accurate as, or more accurate than, the full band (all three bands). The disadvantages of using a full band were especially apparent in the season with high classification accuracy. This study indicates that using all given bands does not necessarily result in the highest classification accuracy. This study also suggests that band selection within the scope of forest type and seasonal variability can contribute to better forest-cover-type classifications.

[1]  Mary E. Martin,et al.  Determining Forest Species Composition Using High Spectral Resolution Remote Sensing Data , 1998 .

[2]  R. Nelson,et al.  Classifying northern forests using Thematic Mapper Simulator data , 1984 .

[3]  T. L. Coleman,et al.  Monitoring forest plantations using Landsat Thematic Mapper data. , 1990 .

[4]  G. Badhwar,et al.  Separability of boreal forest species in the Lake Jennette area, Minnesota , 1985 .

[5]  R. G. Oderwald,et al.  Assessing Landsat classification accuracy using discrete multivariate analysis statistical techniques. , 1983 .

[6]  Jacob Cohen.  A Coefficient of Agreement for Nominal Scales , 1960 .

[7]  Lorenzo Bruzzone,et al.  A technique for feature selection in multiclass problems , 2000 .

[8]  N. Ishitsuka,et al.  Crop discrimination with multitemporal SPOT/HRV data in the Saga Plains, Japan , 2001 .

[9]  Carl W. Ramm,et al.  Correct Formation of the Kappa Coefficient of Agreement , 1987 .

[10]  R. Congalton,et al.  Evaluating seasonal variability as an aid to cover-type mapping from Landsat Thematic Mapper data in the Northeast , 1995 .

[11]  S. Franklin.  Remote Sensing for Sustainable Forest Management , 2001 .

[12]  A. Jones,et al.  The Land Cover Map of Great Britain: an automated classification of Landsat Thematic Mapper data , 1994 .

[13]  Russell G. Congalton,et al.  A review of assessing the accuracy of classifications of remotely sensed data , 1991 .

[14]  R. W. Fitzgerald,et al.  Assessing the classification accuracy of multisource remote sensing data , 1994 .

[15]  Sotaro Tanaka,et al.  Improvement of forest type classification by SPOT HRV with 20 m mesh DTM. , 1990 .

[16]  J. E. Pinder,et al.  A comparison of Landsat Thematic Mapper and SPOT multi-spectral imagery for the classification of shrub and meadow vegetation in Northern California, U.S.A. , 1997 .

[17]  M. Nilsson,et al.  Applications using estimates of forest parameters derived from satellite and forest inventory data , 2002 .

[18]  Darrel L. Williams,et al.  Use of Remotely Sensed Data for Assessing Forest Stand Conditions in the Eastern United States , 1986, IEEE Transactions on Geoscience and Remote Sensing.

[19]  Richard G. Oderwald,et al.  Spectral Separability among Six Southern Tree Species , 2000 .