Stratification approach for forest cover type and landuse mapping using IRS-1A LISS-II data — A case study

Indian Remote Sensing Satellite-1A (IRS-1A) LISS-II data of 24th Nov., 1988 was analysed digitally to differentiate three density classes viz. dense/closed forest, open forest and degraded forest within each vegetation type in the district, Jalpaiguri, West Bengal. Stratification approach was used to classify separately forest cover into pure sal forests, mixed forests, riverine forests along with man-made sal/teak plantations. In this approach the forested and non-forested areas were classified separately through supervised classification techniques using maximum likelihood algorithm using VAX 11/780 based VIPS-32 Image Processing software. Later the two classified outputs were composited to provide entire area of the district. The forest cover of the district was 1420.89 sq. km, (22.82 percent). Other broad landuse/landcover dominant in the district include agricultural areas.(45.20 percent) and tea gardens (10.49 percent). The accuracy of the classified output was estimated to be 90 percent for forested areas and 85 percent in case of other landuse/landcover classes.     

Evaluation of microwave remote sensing data for forest stratification and canopy characterisation

The study presents digital preprocessing techniques, visual mapping capability of airborne X-band SAR data having diverse vegetation types in tropical wet climate. Spatial textural analysis methods have also been evaluated to enhance discriminability of the forest types and features. Attempt has been made to enhance the information by merging optical remote sensing data with microwave X-band response. Finally, the backscattering digital values have been correlated with qualitative and quantitative vegetation parameters. The Leaf Area Index has shown significant relationship with SAR image digital number value.     

Procedures for computation of saturation radiances: A case study for proposed sensors onboard Indian remote sensing satellite-1C

Saturation radiance (SR) is one of the important parameters required while designing a sensor and it depends on application requirements as well as sensor requirements. This study describes the procedures for computation of SR values using theoretical calculations of earthatmosphere radiances as well as radiances observed by existing satellites such as IRS-1A/1B LISS-I, Landsat TM and SPOT sensors. A possible set of saturation radiances for IRS-1C LISS-III, WIFS and PAN sensors has been suggested.     

Crop acreage estimation using ERS-1 SAR data

Synthetic Aperture Radar (SAR) data from the European Remote Sensing Satellite (ERS-1) acquired in July, October and November, 1992 covering the kharif season of the region were used separately and in combination to identify the major crops and for estimation of their acreage before harvest Separability indices were calculated for major cover types and it was found that single-date SAR data cannot be used for accurate identification of various crops. Multi-temporal colour composite facilitated better identification of crop types. Comparison of area estimates made with ERS-1 SAR and IRS-1B LISS II data showed that the commonly used digital data analysis techniques (per pixel classifiers) are not adequate for accurate estimation of crop acreage using SAR data.     

Potential of satellite data in delineation of wastelands and correlation with ground information

In the present study an attempt has been made to delineate the wastelands at microlevel in Matar taluka of Kheda district. IRS OSS II digital data of 1991 has been used for the study which was analysed on VAX 11/780 image processing system. The digital data was classified following supervised classification algorithm. The three types of wastelands namely waterlogged, pasture/grazing and salt affected lands could be identified and mapped. The area statistics for the different wasteland categories present in the taluka were generated. Chemical analysis was also carried out for soils collected from different categories of wastelands to understand the exact nature of soils of each category.     

Evaluation of Cartographic accuracy of IRS-1A LISS II data for upgrading linear details on the existing maps (A case study of Chandigarh area)

Considering the fest pace of development, up-to-date maps have become imperative in developmental planning. Conventional map updating techniques are expensive and time consuming. The present paper makes an attempt to tackle the problem using GIS techniques. Maps generated from IRS-1A LISS II data and Survey of India (SOI) toposheets were used as input maps in GIS. These maps were overlaid to obtain the positional errors at road junction points. The area statistics of each sector were calculated taking the SOI map as reference data. The results show that the areas calculated from LISS II data and SOI map are well comparable and the deviations are within 10%, whereas the positional accuracy (MSE) of points is within 30 m. The study demonstrates that the IRS-1A LISS II data can be used for updating of maps on scale 1:100,000 and smaller in areas where linear features are identifiable in the IRS LISS II image.     

Planform analysis of Hooghly river metamorphosis using multidate satellite data

Conventional techniques of monitoring both long-term and short-term changes in river morphology is time consuming and expensive. Remote sensing with its synoptic and temporal capability is an effective tool to study river metamorphosis. The present study is carried out in lower Ganga basin where Hooghly river is an important river contributing sediment to this vast fertile plain. A time series planform analysis is carried out between 1971 to 1991 period. Thematic maps generated from Landsat MSS and IRS-IA LLSS-I satellite data for 1986 and 1991 period respectively is compared with 1971 topographic maps. The migratory character of paleocourse of Hooghly river is deciphered through the paleochannels, oxbow takes and meander scars deciphered from satellite imagery. The shortterm river metamorphosis like cut-offs, bar migration and point bar accretion morphology are mapped and compared with 1971 topographic maps. The critical zone with respect to cut-offs and future channel migration are identified for river stabilization work.     

A package for digital processing and automatic contouring of aeromagnetic data

Aeromagnetic data processing refers to the processing of magnetic data acquired through airborne surveys. The total magnetic field at various points along the predetermined flight path is recorded using the on board cesium magnetometer. At each of these points, the positional information in terms of latitude, longitude obtained from Inertial Navigation System (INS) and attitude, time and date are also recorded. The track flown by the aircraft while data acquisition is also recorded using video flight path camera. A software package has been developed to process this data and generate outputs in a form amenable to visual interpretation. It is developed on a VAX 11/780 system using VAX FORTRAN and PLOT 10 GKS software. This paper gives an outline of the various functions available in this package and also highlights the problems encountered in choosing the most appropriate techniques/methods to be used in the gridding and contouring of the data.     

Area estimation of arecanut plantations in Sirsi Taluk using IRS data

Extent and distribution pattern of arecanut plantations in Sirsi Taluk, Uttara Kannada District, Kamataka have been studied using IRS, LISS II data. The plantations are found to exhibit perfect zonality, distinct structure and contrasting tonal characteristics and thereby enable their differentiation from other land use/land cover categories encountered in the area. The study not only established the utility of IRS data in acreage estimation of this unique category of land use having significant economic relevance in the area but also in assessing the scope for planning development of spices in these plantations.     

Spectral reflectance characteristics of salt-affected arid soils of Rajasthan

The spectral reflectance characteristics of different types of natural and anthropogenic salt-affected soils have been studied under field conditions. The spectral reflectance value for non-saline and all types of salt-affected soils was maximum in near infra red region (800–1000 nm). The natural salt-affected soils having surface salt encrustation showed highest reflectance value followed by the sodic soils (formed due to high residual sodium carbonate water irrigation) natural saline soils and saline soils due to saline water irrigation. Soil texture, pH, CaC0₃ and organic matter together accounted for 29.6% variation in the maximum reflectance percentage value out of which only pH accounted for more than half (14.2% variation).     

Delineation of social forestry plantations under various afforestation programmes using satellite digital data

Digital analysis of IRS 1A LISS-II (23-03-1991) data was used to delineate social forestry plantations in Matar taluka of Kheda district, Gujarat. VAX 11/780 image processing system was used to categorise the various plantations using supervised maximum likelihood classification. The classified output yielded three categories of plantations viz.,Eucalyptus spp.,Acacia spp., and mixed spp. The statistics for all the categories were also generated.     

Change detection study of islands in Hooghly estuary using multidate satellite images

River estuarine environment constitutes a highly dynamic fluvio-morphological setting where processes of accretion and deposition are active. Hooghly estuary, being one of the largest of estuaries in the east coast of India, needs constant monitoring. Multidate satellite images of IRS-1A L1SS-I and Landsat MSS for 1975–1991 period are studied to detect long term morphological changes in this estuary. The study reveals that the estuarine islands like Sagar, Ghorarmara and Suparbhanga are eroding whereas Lohachara islands has completely eroded off Nayachara island near Haldia due to its shape and size bifurcates the river into two channels. The island as revealed from Satellite images is in accretional phase where the total surface area has increased. The study, therefore, indicates that constant monitoring of spatial and temporal changes in this type of environment would help to understand physical processes.     

Geomorphic evolution of a part of Middle Ganga Plain,Bihar: A case study using remote sensing data

An insight into the geomorphic evolution of any area can be obtained by detailed landform mapping. In the present study, an area in the Middle Ganga Plain has been selected for the study using mainly remote sensing data. Various fluvial landforms have been mapped and the changes in planform of rivers over approximately 50 years have been evaluated. Both fluvial processes and tectonic activities are considered to have collectively influenced the migration of the rivers in this region. Digital enhancements of Landsat MSS and TM data are found to be quite useful in identification and mapping of subtle fluvial palaeofeatures. The present study demonstrates the utility of remote sensing in examining the geomorphic evolution of the area.     

Ranchi mega lineament and its correlation with geological and geophysical data

Remotely-sensed satellite data of Landsat 5 (TM) and IRS-1A (LISS II) covering parts of central portion of Chotanagapur plateau have been analysed visually to delineate some prominent lineaments in Ranchi district and its adjoining area. One of the most prominent lineament “Ranchi mega lineament? has been analysed in this paper. This lineament is a potential site for base metal mineralisation at some places, and it acts as good groundwater conduit in the region. No earlier attempt to map this entire structural unit as a separate entity is known. Geophysical bouguer gravity anomaly map, lithological and structural maps have some remarkable correlation with this lineament. In the present paper authors have tried to analyse the lineament with respect to its tectonic, mineralisation and groundwater prospects. Resistivity survey carried in close proximity of this lineament at selected sites, indicate that rocks are well fractured at depth. Available lithological and structural map support this lineament as a shear near Purulia (W.B.) and also having base-metal mineralisation potential in this area. Authors opine that existing gravity anomaly map may be further augmented in the light of present study.     

Assessment of sediment distribution pattern in the Tungabhadra reservoir using satellite imagery

In order to understand the dynamic aspects of suspended sediments in an inland water body, Tungabhadra reservoir on the Tungabhadra river in the Krishna basin was studied. The study has been carried out using Landsat MSS and IRS-1 A LISS-II images. Visual interpretation techniques have been used to obtain information on the location and extent of sediment distribution pattern in the water-spread area of the reservoir. It has been possible to monitor the seasonal fluctuations in the reservoir water-spread; measure corresponding fluctuation in the volume of water in the reservoir, and study seasonal changes in the suspended sediment distribution pattern in the reservoir. An attempt has also been made to prepare area capacity curve for the reservoir. Semi-quantitative assessment of sediment deposits between reservoir levels were made considering water spread area from the satellite images (May 1986, April 1987, Jan. 1988, Jan. 1989 and March 1989) and sedimentation survey report of KERS 1985, (Karnataka Engineering Research Station). The results indicated that the high Concentration of sediments is at the western confluence of the Tungabhadra river. On the basis of tonal variation as observed, the reservoir could be divided into four major zones, viz., very high and high at the river confluence, moderate at the periphery and low at the dam site.     

Assessment of soil degradation Hazards in Jalor and Ahor Tehsil of Jalor district (Western Rajasthan) by remote sensing

In the present study efforts have been made to identify and map areas affected by various land degradation processes with the aid of Landsat TM imagery data of 1988 and ground truth verification. The kind, extent and degree of land degradation have been mapped. In an area of over 4,124 sq. km. 51% was affected by water erosion and 30% area by wind erosion. Nearly 1.14% area is affected by salinity. Degradation due to combined effect of water and wind erosion and water erosion and salinization has affected 8.20% of the study area. 1.53% area is free from any hazard. Remaining 7.85% area comes under hills and rivers. Nearly 44 percent of the affected area is subjected to moderate and severe degradation which can easily be combatted by techniques referred.     

Differentiation of oil seed crops by remote sensing based on their spectral reflectance pattern

The feasibility of differentiating four oil-seed crops viz., mustard, toria, yellow sarson and sunflower, based on their spectral reflectance in the visible and near infra red region was studied in a field experiment, The spectral vegetative index profiles, generated during the growth period of different oil seed crops indicated two vegetative growth peaks and a depression between the two peaks, due to the conspicuous yellow colour of flowers, which masked the green leaves. The magnitude of such depression in the spectral vegetation indices viz., ‘Greenness’ and ‘Perpendicular Vegetation Index’ (PVI), were of higher magnitude in yellow sarson. The flowering period parameters viz., flowering time, duration and intensity, deduced from the spectral vegetation indices were found to be beneficial in differentiating different oil-seed crops by remote sensing. A plot of ‘Brightness’ vs. ‘Greenness’ values determined during the growth of the crops formed typical clusters. The cluster representing toria crop was significantly different from the other crops, thereby making toria identifiable from others by remote sensing.     

On the equivalence of spherical splines with least-squares collocation and Stokes’s formula for regional geoid computation

This work is an investigation of three methods for regional geoid computation: Stokes’s formula, least-squares collocation (LSC), and spherical radial base functions (RBFs) using the spline kernel (SK). It is a first attempt to compare the three methods theoretically and numerically in a unified framework. While Stokes integration and LSC may be regarded as classic methods for regional geoid computation, RBFs may still be regarded as a modern approach. All methods are theoretically equal when applied globally, and we therefore expect them to give comparable results in regional applications. However, it has been shown by de Min (Bull Géod 69:223–232, 1995. doi: 10.1007/BF00806734) that the equivalence of Stokes’s formula and LSC does not hold in regional applications without modifying the cross-covariance function. In order to make all methods comparable in regional applications, the corresponding modification has been introduced also in the SK. Ultimately, we present numerical examples comparing Stokes’s formula, LSC, and SKs in a closed-loop environment using synthetic noise-free data, to verify their equivalence. All agree on the millimeter level.     

Assessing the Priorities for Sustainable Forest Management in the Sikkim Himalaya,India: A Remote Sensing Based Approach

Sikkim is a small, mountainous, Indian state (7,096 km²) located in the eastern Himalayan region. Though a global biodiversity hotspot, it has been relatively less studied. A detailed forest type, density and change dynamics study was undertaken, using SATELLITE remote sensing data and intensive field verification. The landscape was found to be dominated by alpine and nival ecosystems, with a large portion above the tree line, considerable snow cover, and a sizeable area under forest cover (72%, 5,094 km²). A total of 18 landscape components including 14 vegetation classes were delineated, with the major ones being oak forest, alpine meadow, alpine scrub, conifer forest and alder-cardamom agro-forestry. Of the 3,154 km² of forests below the tree line, 40% were found to be dense (>40% tree canopy cover). A sizeable portion of the non dense forests below the tree line was contributed by the degradation of oak forests, which was confirmed by change detection analysis. However on a positive front over the past decade, ban on grazing and felling of trees in forests has been implemented. In order to expand the extent of dense forests, further efforts are needed for the restoration of oak forests such as fire protection, providing alternatives to firewood use, promotion of alder-cardamom agro-forestry in the private lands and protection of the small-sized, fragmented forest patches in the subtropical belt.     

Quantification of carbon stock to understand two different forest management regimes in Kayar Khola watershed,Chitwan, Nepal

The impact of forest management activities on the ability of forest ecosystems to sequester and store atmospheric carbon is of increasing scientific and social concern. This is because a quantitative understanding of how forest management enhances carbon storage is lacking in most forest management regimes. In this paper two forest regimes, government and community-managed, in Kayar Khola watershed, Chitwan, Nepal were evaluated based on field data, very high resolution (VHR) GeoEye-1 satellite image and airborne LiDAR data. Individual tree crowns were generated using multi-resolution segmentation, which was followed by two tree species classification (Shorea robusta and Other species). Species allometric equations were used in both forest regimes for above ground biomass (AGB) estimation, mapping and comparison. The image objects generated were classified per species and resulted in 70 and 82 % accuracy for community and government forests, respectively. Development of the relationship between crown projection area (CPA), height, and AGB resulted in accuracies of R² range from 0.62 to 0.81, and RMSE range from 10 to 25 % for Shorea robusta and other species respectively. The average carbon stock was found to be 244 and 140 tC/ha for community and government forests respectively. The synergistic use of optical and LiDAR data has been successful in this study in understanding the forest management systems.