Symbolic agglomerative clustering for quantitative analysis of remotely sensed data

The launch of the first Indian Remote Sensing Satellite in 1988 ushered in a new era of utilization of satellite data and technology in natural resources monitoring, coastal area management, disaster mitigation, weather and climate change studies. India with its very long coast line, areal extent, resources and population would have been a huge challenge to the coastal researchers and administrators if not for the Indian Remote Sensing satellite data that provided the much needed temporal, thematic information for effective mapping, monitoring and management of the coastal areas. The country’s large and diverse coastal regions are economically important and ecologically rich but also extremely vulnerable and highly prone to natural and anthropogenic hazards. Through the last three decades, the Indian remote sensing programme has contributed immensely to our understanding of the coast and ocean processes, coastal ecosystems, habitats, marine pollution and the vulnerability of the coast to natural hazards. IRS-1C with its improved spatial resolutions, additional spectral band, improved repetitively ushered in a new era of research specifically in the management of coastal areas and resources. Availability of spatial data at various resolutions, low cost, advances in technology resulted in new applications for management coastal areas. IRS-1C also can be aptly termed as the precursor for thematic satellite series such as OCEANSAT, RESOURCESAT and CARTOSAT, thereby setting the environs for thematic research in coastal areas. Integrated utilization of remote sensing data, numerical models, GIS and GPS has paved the way for the development of operation systems for management of coastal areas, thereby realizing the vision of Dr. Vikram Sarabhai, the father of Indian Space Programme, of building a self-reliant and indigenous Space Programme for the betterment of quality of life of the common people.

IR Earth horizon sensors are used for attitude determination of satellites. Simulation of IR horizon and ground calibration are major problems for testing sensors used in low altitude orbits due to geometrical constraints. A large earth simulator and a test facility with a 3500 mm. diameter disc have been developed for the calibration and the qualification of different types of IR earth sensors used on-board Indian Remote Sensing (IRS) satellites. The unique design of the test facility enables the simulation of the satellite altitudes, roll and pitch angles, IR radiation from space and Earth, gradients due to seasonal changes in radiance and thermo-vacuum environments of the sensor under test. The sensor performance can be tested either in open air conditions or under clean vacuum conditions at different temperatures of the sensor. The gradient of the Earth's IR radiation which changes with latitudes and seasons, and the Sun and the Moon interference can also be simulated. A computer based data acquisition system along with the instrumentation is used for controlling and monitoring vacuum, linear and rotary motions, temperature distribution on the earth disc and also for monitoring and recording of the sensor outputs. This paper describes the methodology behind various simulations and the testing involved. The constructional details of the large earth simulator and the alignment of the sensor with respect to the earth simulator are explained. A few of the actual sensor test results obtained are also included.

Identification of homogeneous regions in a satellite image is essentially the clustering of pixels in intensity space. Importantly remote sensing image like satellite images contain varieties of land cover types. Some of the covers are significantly large areas whereas some are relatively smaller regions. Therefore, automatically detecting such wide varying areas is a challenging task. Hence, this fact motivated us to propose an improved clustering technique viz. Ensemble based Differential Evolution for Fuzzy Clustering (EDEFC). For this purpose, very recently developed three variants of differential evolution (DE) are used in order to perform the clustering with different set of solutions. As a result, better clustering solution yields from the ensemble of DEs by exhaustive exploration of search space. The proposed EDEFC technique is applied on two numeric remote sensing datasets and Indian Remote Sensing (IRS) satellite image of Kolkata. The results of the EDEFC are shown quantitatively and visually by comparing with eight other clustering techniques. Moreover, the statistical significance test has also been performed in order to judge the superiority of EDEFC.

Abstract. Remote Sensing along with Geographical Information System (GIS) has been proven as a very important tools for the monitoring of the Earth resources and the detection of its temporal variations. A variety of operational National applications in the fields of Crop yield estimation , flood monitoring, forest fire detection, landslide and land cover variations were shown in the last 25 years using the Remote Sensing data. The technology has proven very useful for risk management like by mapping of flood inundated areas identifying of escape routes and for identifying the locations of temporary housing or a-posteriori evaluation of damaged areas etc. The demand and need for Remote Sensing satellite data for such applications has increased tremendously. This can be attributed to the technology adaptation and also the happening of disasters due to the global climate changes or the urbanization. However, the real-time utilization of remote sensing data for emergency situations is still a difficult task because of the lack of a dedicated system (constellation) of satellites providing a day-to-day revisit of any area on the globe. The need of the day is to provide satellite data with the shortest delay. Tasking the satellite to product dissemination to the user is to be done in few hours. Indian Remote Sensing satellites with a range of resolutions from 1 km to 1 m has been supporting disasters both National & International. In this paper, an attempt has been made to describe the expected performance and limitations of the Indian Remote Sensing Satellites available for risk management applications, as well as an analysis of future systems Cartosat-2D, 2E ,Resourcesat-2R &RISAT-1A. This paper also attempts to describe the criteria of satellite selection for programming for the purpose of risk management with a special emphasis on planning RISAT-1(SAR sensor).

Drainage morphometry and its influence on landform characteristics in a basaltic terrain, Central India – a remote sensing and GIS approach

Indian Remote Sensing satellite (IRS)-1B, Linear Imaging Self Scanner (LISS)-II spectral digital data was analysed to determine the feasibility of quantifying the concentration of suspended solids in the surface water of inland water body, Dal lake, in Srinagar, India. The water samples collected in concurrent with IRS-1B overpass, were analysed to determine the concentration of suspended solids. The results indicate that a positive functional relationship exist between the concentration of suspended solids and the visible wave length bands 1 and 3 and near infrared band 4. It has been observed that as the concentration of suspended solids increase, the spectral response also increases. It is concluded that IRS LISS-H data can be effectively used to quantify suspended sediment concentration in the Dal lake surface water.

This paper reports the bandpass solar exoatmospheric irradiance and Rayleigh optical thickness for six categories of sensors, namely, the linear imaging self scanning sensor (LISS)-I, -II, -III, the wide field sensor (WiFS), the panchromatic (PAN), and the ocean color monitor (OCM) which have been flown on Indian Remote Sensing Satellites (IRS)-1B, -1C, -1D, and -P4 platforms, based on laboratory measurements of relative spectral response (RSR) of each band. These estimates have been compared with similar contemporary international sensors.

Alluvial/valley fills potentially form excellent groundwater zones and therefore their distribution over a terrain is very important in hydrogeological terms. Given the advantage of synoptical view and multispectral characteristics, digitally recorded remotely sensed images and their processing techniques provide a unique opportunity to map such lithounits with great accuracy and swiftness. The study area suffers from acute groundwater deficiency and therefore mapping the alluvial fill zones can help considerably in groundwater exploration and targeting. In this study an attempt was made to develop a comprehensive digital enhancement package to digitally enhance and extract the alluvial/valley‐fill regions from an Indian Remote Sensing Satellite (IRS) 1D Linear Image Self Scanning (LISS) III image. Of the several image enhancement techniques tried, the Hybrid Colour Composite (HCC) and Square root of Multiplied Bands (SMB) methods were found to be particularly effective and best suited for such an exercise. The final alluvial fill map prepared was ground checked through several field surveys, which also included checking the potential groundwater well yield of the mapped zones. The study demonstrates the usefulness of digital image processing (DIP) in mapping alluvial fill zones and the efficacy of remote sensing in groundwater exploration.

The paper presents a geospatial modeling approach for the assessment of plant richness in Barsey Rhododendron Sanctuary in Sikkim, a Himalayan State of India located in the “Indo‐Burma” biodiversity hotspot. Remotely sensed data from Indian Remote Sensing Satellite IRS‐1C Linear Imaging Self‐Scanner (LISS‐III) and field‐based methods were synergistically used to model plant richness on 1:50,000 scale. It was found that the sanctuary is dominated by East Himalayan Moist Temperate Forest (55.50%), followed by Rhododendron Forest (23.77%), Degraded Forest (6.66%) and Hemlock Forest (0.78%). The vegetation map prepared through digital interpretation of satellite imagery was subjected to landscape analysis and assessment of biotic disturbance in terms of disturbance index. The disturbance index together with species richness, ecosystem uniqueness, total importance value and terrain complexity was modeled to assess the plant richness in this unique sanctuary. Out of the 120 km2 of the total geographical area of the sanctuary, 28.45 per cent was found to possess very high plant richness followed by high (50.84%), medium (6.96%) and low richness (13.75%). It was noted that plant richness assessment at ecosystem level presents a more realistic picture than at landscape level. The study demonstrated that remote sensing coupled with landscape analysis, ground inventory data and geospatial modeling holds good potential for rapid and operational assessment of plant richness.

Spherical and oblate Earth conical shadow models for LEO satellites: Applications and comparisons with real time data and STK to IRS satellites

A nonparametric, hierarchical, divisive clustering algorithm based on the mutual nearest neighborhood concept is proposed for classification of large volume of multispectral data indenting minimum memory and computational time. To realise the reduction of memory and computational time requirements, the authors define an area of interest around each pixel to search the required k nearest neighbors necessary for initiating splitting operation. The authors have proposed a modified distance measure, which incorporates the relative position of the pixels together with their feature values in order to improve the performance of the algorithm and reduce the classification error rate. The algorithm is simple, noniterative and can be successfully applied on any image data. Simulation experiments are conducted to corroborate the efficacy of the proposed methodology in terms of time, space and error rate.

Coastal ecosystem is complex, sensitive and is subject to the influences of nature as well as human interferences. In recent years the prawn farming has increased significantly in the coastal areas of Andhra Pradesh, India, due to great demand in the export market at the coast of agricultural areas and mangrove forests which are essential for the coastal ecological balance. Hence, a study was carried out with the objective of mapping and monitoring the area under prawn farming and mangroves using multitemporal remotely sensed data along Krishna delta of Guntur and Krishna districts of Andhra Pradesh. Remotely sensed data from Landsat‐MSS pertaining to 1973, 1985 and 1990 years and Indian Remote Sensing Satellite‐1A (IRS‐1A) of 1992 are used for the study. The methodology adopted is digital analysis of multitemporal data. The study revealed that during 1985 to 1992 period the area under prawn farming increased significantly. Initially prawn cultivation started on a smaller scale on saline waste lands along the coast. Later due to commercialization of prawn culture, its cultivation has been extended into agricultural lands and mangrove forest.

Neural networks are growing in popularity today as a tool for classification of remotely sensed images. One of the major stumbling blocks for neural networks to be accepted for operational use in remote sensing has been the long computational times involved. This is particularly relevant in the case of the well-known back-propagation (BP) algorithm trained multilayer perceptron (MLP) classifier. The long time involved in training the network to realize the relationship between the training data and their assigned classes is due to lack of a sound theory to set the initial network parameters. Errors in classification are partly due to mixed pixels and absence of contextual information, which is almost always used by human interpreters while analyzing images. An approach is suggested here to take advantage of global optimization techniques to assist in proper training of the MLP classifier, and use of relaxation labelling algorithms to refine the MLP output based on the contextual information present within small local neighbourhoods of pixels. The results are illustrated using Indian Remote Sensing (IRS) images.

Land-use information is required for a number of purposes such as to address food security issues, to ensure the sustainable use of natural resources and to support decisions regarding food trade and crop insurance. Suitable land-use maps often either do not exist or are not readily available. This article presents a novel method to compile spatial and temporal land-use data sets using multi-temporal remote sensing in combination with existing data sources. Satellite Pour l'Observation de la Terre (SPOT)-Vegetation 10-day composite normalized difference vegetation index (NDVI) images (1998–2002) at 1km2 resolution for a part of the Nizamabad district, Andhra Pradesh, India, were linked with available crop calendars and information about cropping patterns. The NDVI images were used to stratify the study area into map units represented by 11 distinct NDVI classes. These were then related to an existing land-cover map compiled from high resolution Indian Remote Sensing (IRS)-images (Liss-III on IRS-1C), reported crop areas by sub-district and practised crop calendar information. This resulted in an improved map containing baseline information on both land cover and land use. It is concluded that each defined NDVI class represents a varying but distinct mix of land-cover classes and that the existing land-cover map consists of too many detailed ‘year-specific’ features. Four groups of the NDVI classes present in agricultural areas match well with four categories of practised crop calendars. Differences within a group of NDVI classes reveal area specific variations in cropping intensities. The remaining groups of NDVI classes represent other land-cover complexes. The method illustrated in this article has the potential to be incorporated into remote sensing and Geographical Information System (GIS)-based drought monitoring systems.

Water depth estimation using optical remote sensing offers a reliable and efficient means of mapping coastal zones. Here, we aim to find a suitable model for fast and practical bathymetry of an estuary using Indian Remote Sensing Satellite (IRS) Linear Imaging Self Scanning Sensor (LISS-3) images. The study examines three different models; (1) least square regression model, (2) spectral band-ratio method and (3) multi-tidal bathymetry model. The findings are supported with in situ observed depth values and statistical estimates. Although the least square regression model has provided best results with root mean square error (RMSE) of 0.4 m, it requires a large number of observed data points for absolute depth estimation. Spectral band-ratio and multi-tidal model provides results with RMSEs 2.1 and 0.9 m, respectively. The present investigation demonstrates that multi-date imagery exploitation at disparate tide levels is the best estimation technique for recursive shallow water bathymetry where in situ observation is not possible.