THEMATIC MAPPING USING QUICKBIRD MULTISPECTRAL IMAGERY IN OUNG EL-JEMEL AREA, TOZEUR (SW TUNISIA) Belgium

Transcription

1 THEMATIC MAPPING USING QUICKBIRD MULTISPECTRAL IMAGERY IN OUNG EL-JEMEL AREA, TOZEUR (SW TUNISIA) FRANCESCO G. DESSÌ 1 AND ABDOUL JELIL NIANG 2 1 Earth Science Department, TeleGis Lab, Via Trentino 51, Cagliari, Italy fdessi@unica.it 2 Laboratoire de Géomorphologie et Télédétection, Allées du 6 Aout 2, B Liège, Belgium ajeniang@yahoo.fr Abstract Standard ground methods of land use mapping are labor intensive, time consuming and are done relatively rarely. In recent years, satellite remote sensing techniques have been developed, with increasing availability of very high resolution satellite imagery, which have proved to be of immense importance for preparing accurate land use and land cover maps. In case of inaccessible region, this technique is perhaps the only method of obtaining the necessary data on a cost and time-effective basis. The aim of this work was to investigate the use of Quick Bird imagery for thematic mapping in Oung El-Jemel area, close to the town of Tozeur (SW Tunisia), supported by field verifications data. Keywords: Thematic mapping, Quick Bird, NDVI 1 Introduction The Oung El-Jemel area is located in SW Tunisia (Fig. 1) and takes its name from a unique rock formation near the salt lake Chot el Gharsa. This area has been famous among locals for its camel-like shape and recently has become a tourist attraction following a number of movies being filmed there. The principal process is aeolian, with a variety of mobile sand dunes and semifixed sand dunes. The moving sand is also forming erosion landforms, such as yardangs. In addition to these aeolian processes, fluvial erosion is forming small gullies along the shoreline of the Chott el Rhasa, which consists of exposures of Villafranchian sediments. 2 Quick Bird satellite data Quick Bird offers panchromatic and multispectral imagery with the highest spatial resolution currently available within the satellite sensors. The Quick Bird satellite has panchromatic and multispectral sensors with resolutions at cm and m, A. Marini and M. Talbi (eds.), Desertification and Risk Analysis Using High and Medium Resolution Satellite Data, Springer Science + Business Media B.V

2 208 F.G. Dessì and A.J. Niang Fig. 1: Study area and camel rock respectively, depending upon the off-nadir viewing angle (0 25 ). The sensor has coverage of 16.5e19 km in the across-track direction. In addition, the along-track and across-track capabilities provide a good stereo geometry and a high revisit frequency of days depending on latitude. Quick Bird technical notes are available on line at the web site The panchromatic sensor collects information at the visible and near infrared wavelengths and has a bandwidth of nm. The multispectral sensor acquires data in four spectral bands from blue to near infrared (NIR). Both panchromatic and multispectral sensors offer 11 bit (2,048 grey levels) resolution. The Quick Bird imagery products are available at different processing levels (basic, standard, and ortho) serving the needs of different users. In this work we have used a Quick Bird image acquired under clear sky conditions on 18th February Methodology 3.1 Satellite data processing The idea that lies behind our analysis is that features could be enhanced and efficiently studied by exploiting the high spatial resolution of satellite Quick Bird panchromatic data and the multispectral properties of the four spectral channels. The methodological approach (shown in Fig. 2) adopted for the enhancement and extraction of homogenous areas is mainly based on photointerpratation (Richardson and Wiegand, 1977; Myneni et al., 1995). Data fusion refers the process of combining multiple images of a scene to obtain a single composite image. The different images to be fused can come from different sensors of the same basic type or they may come from different types of sensors. The composite image should contain a description of the scene more useful than those provided by any of the individual source images. In the current cases under investigations the Quick Bird panchromatic and multispectral images were fused (Zhang, 2004) by using pan-sharpening in ENVI software.

3 Thematic Mapping Using Quickbird Multispectral Imagery in Oung El-Jemel Area, Tozeur 209 Fig. 2: Flow chart of the methodology adopted 3.2 NDVI Additionally, on the basis of remotely sensed data, vegetation be suitably identified by exploiting vegetation indexes (Bannari et al., 1995; Elvidge and Chen, 1995) that are spectral combinations of different bands. Such indexes are quantitative measures, based on vegetation spectral properties, that attempt to measure biomass or vegetative vigour. Vegetation indexes are mainly derived from reflectance data from the Red and near infrared (NIR) bands. They operate by contrasting intense chlorophyll pigment absorption in the red against the high reflectance of leaf mesophyll in the near infrared. The simplest form of vegetation index is the ratio between two digital values from the red and near infrared spectral bands. The most widely known and used ratio-based index is the normalized difference vegetation index (NDVI) (Rouse et al., 1974; Carlson and Ripley, 1997). The NDVI is calculated as the ratio of the difference between the near infrared and the red band and the sum of the two bands by using the following formula: NDVI = (NIR - Red) / (NIR + Red) The normalization of the NDVI reduces the effects of variations caused by atmospheric contaminations. NDVI is indicative of plant photosynthetic activity and has been discovered to be related to the green leaf area index and the fraction of photo-synthetically active radiation absorbed by vegetation. Using this index, the difference between vegetation and non-vegetation is emphasized. Reflectance values for vegetation have their maximum in the near infrared and a minimum in the red spectral domain. High values of the NDVI indicate vegetation and values around 0 non-vegetated land areas (Figs. 3, 4 and 5).

4 210 F.G. Dessì and A.J. Niang Fig. 3: Homogeneous domains in the Oung El-Jemel area Fig. 4: NDVI for a portion of the Quick Bird image of the study site. Bright values indicate high level of vegetation and dark values non-vegetated areas

5 Thematic Mapping Using Quickbird Multispectral Imagery in Oung El-Jemel Area, Tozeur 211 Fig. 5: Portion of Quick Bird scene of the study area. In the southern zone some mobile sand dunes. In the northern area some semi-fixed dunes and sporadic vegetation. FCC 432 with Pan-sharpening Chott It is the border of a sebkha covered with halophilic vegetation; a sebkha being an ephemeral salt water body Sebkha A geologic feature, in North Africa, which is a smooth, flat, plain usually high in salt; after a rain the plain may become a marsh or a shallow lake until the water evaporates Clayey soil rangeland Clayey soil rangelands extend along the border of the Chott and are covered with halophilic vegetation. Salty surface crusts are frequently present in this area Rangelands on vegetated Nebkhas Correspond to rangeland on semi-mobile dunes. The vegetation consists of small (<1 m) bushes that allow V shaped Nebkhas development Human settlements Mainly consist of camps of nomads. Among the few building of the area, we can mention the installations constructed for the movie Stars Wars.

6 212 F.G. Dessì and A.J. Niang Nebkhas Rangelands similar to those on vegetated Nebkhas, but with less vegetation and a higher aeolian activity Compact bare soils Flat area free from sand and with a very small vegetation (Drake, 1997) cover located between the nebkhas and the area of migrating dunes Migrating dune Migrating (or shifting) dunes cover all the area. These barchans extend up to 100 m in width and 10 m in height. This area is free from vegetation cover. 4 Conclusions Satellite Quick Bird imagery were used in order to assess their capability to detect objects in the study area. The use of data fusion and photointerpretation procedures improves the identification of homogeneous domains. The integration of results obtained from panchromatic and image fusion products provides valuable information for a detailed physical and geometrical characterisation of the region. The use of NDVI allowed bettering enhancing surfaces covered by green and herbaceous plants. Bibliography Bannari A., Morin D., Bonn F., Huete A.R A review of vegetation indices, in Remote Sens. Rev. 13, Carlson T.N., Ripley D.A On the relation between NDVI, fractional vegetation cover, and leaf area index, in Remote Sens. Environ. 62, Drake D.A Recent aeolian origin of superficial gypsum crusts in Southern Tunisia: geomorphological, archaeological and remote sensing evidence, in Earth Surf. Proc. Land. 22, Elvidge C.D., Chen Z Comparison of broadband and narrow-band red and near infrared vegetation indices, in Remote Sens. Environ. 54, Myneni R.B., Maggion S., Iaquinta J., Privette J.L., Gobron N., Pinty B., Verstraete M.M., Kimes D.S., Williams D.L Optical remote sensing of vegetation: modelling, caveats and algorithms, in Remote Sens. Environ. 51, Richardson A.J., Wiegand C.L Distinguishing vegetation from soil background information, in Photogramm. Eng. Rem. Sens. 43, Rouse J.W., Haas Jr., R.H., Schell J.A., Deering D.W Monitoring Vegetation Systems in the Great Plains with ERTS, in NASA SP-351, 3rd ERTS-1 Symposium, Washington, DC, pp Zhang Y Understanding image fusion, in Photogramm. Eng. Rem. Sens., 70 (6),