Expression of Qualitative and Quantitative Information on Mangrove Forest in Okinawa
Munetake Kanetomi*, Kazuhiro Sto** and Takahisa Yamauchi*
* : Air-Graph Co. Ltd., Kamiigusa 1749-1, Kawashima,
Hiki, Saitama 350-0152, Japan
Tel: 81-492-97-7410, Fax: 81-492-97-7414
**: College of Agriculture, University of the Ryukus,
Senbaru 1, Nishihara, Okinawa 903-0129, Japan
Tel: 81-98-895-8792, Fax: 81-98-895-8734
Abstract
Qualitative and quantitative simultaneously expressed map ought to be more useful for forest management. The way to put this matter into practice using remote sensing was examined for mangrove forest in Okinawa.
Two kind of information were expressed as imagery. Patterns composed with black marks were used for smaller number of classes and colors were used for larger number of classes. In this paper, qualitative classification as species and mixture ratio was expressed by patterns and quantitative classification as stand stock was done by colors. In this case of that one dot corresponded to one pixel, problem were remained.
Introduction
We have examined qualitative and quantitative simultaneous expression of information on mangrove forest using remote sensing technics4.5. It is assumed that qualitative information are forest type, operation type, species, mixture ratio and so on, and quantitative information are height, diameter at breast height (DBH) and stand stock or biomass. Some classifications are useful for qualitative classification and regression analysis is useful for quantitative estimation. One of the authors and other coresearchers reported relationship among TM data of landsat 5 and stand parameters on mangrove forest
1, 2, 3). Although simultaneous expression with colored qualitative classification and isopleth figure of estimated quantitative stand parameter was suggested in the reports, systematic drawing of the isopleth figure is not completed yet.
We tried to express qualitative classification as patterns composed with black marks. It is easier to make pattern correspond to a qualitative class rather than to quantitative class because number of classes is usually smaller. It is possible to express quantitative classes as considerable many colored classes. So monochrome patterns were overlaid on the colored classification imagery.
Method
A mark can be expressed only one as a point by a dot on a CRT. It becomes easier to make many marks increasing number of dots. A square corresponded to a pixel is composed on a CRT by dots. The size must be same for both information because one is overlaid to another on a CRT. We fixed the size as five dots for a pixel because of clearness of pattern, ease to make marks and extent of background. Simple and clear examples of mark composed with five by five dots were shown in Fig.-1.
Fig.-1 Simple and clear examples of mark composed with5 by 5 dots
For this examination, we used a set of TM data without band 6 of Landsat 5 acquired on 16 May 1994 as a scene of path 115 and row 43. So the square on a CRT corresponded to pixel equivalent to extent of 30m by 30m. Extensive area can be displayed as imagery with smaller squares on a CRT and displayed area become narrower with larger squares. So the size must be decided harmoniously two conditions of clearness and extent displayed imagery.
a) Preparation of data sets
A mask procedure suggested by us
4, 5) was applied for the preparation of data sets of each band. The mask was made with binary imagery of band 4 and 5. The threshold values were decided from CCT counts along boundary between waters and mangrove area or mangrove area and land area. This mark was composed with Zero inside mangrove area and 255 outside it. From this procedure mangrove area was identified. It becomes possible to produce data sets to hold actual data only inside mangrove area by subtraction of this mask between each band data. These procedure data were used for estimation of stand parameters and classification of mangrove species.
b) Estimation of stand parameters as quantitative information
Relationship between TM data of Landsat or many calculated indexes from them and stand parameters has been examined on mangrove forest in Okinawa
1, 2). Additionally similar correlation and regression analysis has been done for mangrove forest in Sumatra, Indonesia3). In these paper, examined stand parameters were tree height, diameter at breast height (DBH), stem volume, crown area stand stock and so on. And it was described that B4, B1/B4, B2/B4, B3/B4, B4/(B3+B1), B4/(B3+B2), B4/(B3+B7), (B4-B1)/(B4+B1), B4-B2)/(B4+B2) and (B4-B3)/(B4+B3) showed significantly high correlation coefficient with some stand parameters. B1, B2, B3, B4, B5 and B7 meaned CCT count of each band of suffixed number.
In this paper, we had quoted following regression equation for estimation of stand stock as quantitative information from the paper mentioned above.
Y = 6.15X - 352.3 (r = 0.878, significant at 0.1 % level)
Where, Y is stand stock (m
3/ha) for each pixel
And X is CCT count of band 4
Estimated values were divided to ten ranges and displayed by a colored square on a CRT, but by a monochrome square in this paper. This procedure and printed map is useful to sum up volume of resources and to know distribution of resources. It is possible to estimate also tree height, DBH, stem volume as quantitative information.
