Dibyendu Dutta

Daily composite wind fields from Oceansat-2 scatterometer

Oceansat-2 scatterometer (OSCAT) is an active microwave sensor, intended to provide ocean surface wind vectors over the global oceans. In the present work, an attempt has been made to generate daily composites of OSCAT Level-3 (L3) wind vectors using Data-Interpolating Variational Analysis (DIVA) method from ascending and descending passes over the Indian Ocean region. This could be useful for operational purposes and in generating value-added products like wind stress and curl of wind stress. The daily composite wind vectors of zonal (U) and meridional (V) components have been validated by comparing with Advanced Scatterometer (ASCAT) and wind from in situ buoys for the year 2012. Wind composites thus generated using DIVA are found to match well with in situ, and ASCAT wind products. Minor deviations are observed with respect to ASCAT wind, which could be attributed to the difference in interpolation techniques used for the two scatterometer products. Given that the repeat period of ASCAT is 5 days and that of OSCAT is only 2 days, OSCAT wind products could be conveniently used for real-time met-ocean studies.

Land use indicators of a watershed in arid region, western Rajasthan using Remote Sensing and GIS

The vegetation dynamics and land use/land cover types of Birantiya Kalan watershed located in the arid tracts of western Rajasthan have been characterized and evaluated using Remote Sensing and Geographical Information System (GIS). The watershed under study falls in the transitional plain of Luni Basin and is characterized by Aravali ranges in the eastern half and vast alluvial plains in the west. The land use/land cover types, as identified are cropland, fallow, forest, land with scrub, land without scrub, sandy area and the water body. Land with scrub occupied maximum area (39% area of the watershed) in 1996 in place of crop land which was dominant (43% of total area) in the year 1988. During eight years period, seasonal fallow land increased significantly and the areal extent of water body decreased to almost half. Vegetation vigour types have been classified into very poor, poor. moderate, good and very good categories. Moderate vigour type reduced from 62 to 27% and poor type increased from 34 to 68% during the period 1988 to 1996. Other vegetation vigour types have not shown any significant changes. To quantify the changes over the years in both vegetation and land use/land cover, weightages have been given to each type and composite values of both vegetation vigour and land use types for 1996 and 1988 have been calculated. It has been observed that the ratio for vegetation vigour has been found to be 0.85 showing that the overall vegetation have not improved after the treatment. The ratio for land use is found to be 1.01, which indicates negligible change in land use.

Segmentation and classification of high spatial resolution images based on Hölder exponents and variance

Abstract Pixel-based or texture-based classification technique individually does not yield an appropriate result in classifying the high spatial resolution remote sensing imagery since it comprises textured and non-textured regions. In this study, Hölder exponents (HE) and variance (VAR) are used together to transform the image for measuring texture. A threshold is derived to segment the transformed image into textured and non-textured regions. Subsequently, the original image is extracted into textured and non-textured regions using this segmented image mask. Afterward, extracted textured region is classified using ISODATA classification algorithm considering HE, VAR, and intensity values of individual pixel of textured region. And extracted non-textured region of the image is classified using ISODATA classification algorithm. In case of non-textured region, HE and VAR value of individual pixel is not considered for classification for significant textural variation is not found among different classes. Consequently, the classified outputs of non-textured and textured regions that are generated independently are merged together to get the final classified image. IKONOS 1 m PAN images are classified using the proposed algorithm, and the classification accuracy is more than 88%.

Band selection in hyperspectral imagery using spatial cluster mean and genetic algorithms

Dimensionality reduction of hyperspectral images is essential for reduction of computational complexity and faster analysis. A novel method for band reduction has been proposed here, which has been adapted from the genetic algorithm (GA) along with spatial clustering. Spatial clustering generates overall signature variation present in a particular scene and in turn removes huge redundancy present in the raster data set. GA is applied on the clustered signatures to extract the reduced set of bands that is computed to be the “fittest” i.e., those bands that provide the most discriminating information in a hyperspectral image. This has been computed by taking the sum of Kullback–Leibler divergences (KLD) between consecutive selected bands. A higher KLD value amongst adjacent selected band implies higher divergence in value. The selected band-set image has been classified and the accuracy indices are evaluated respectively. The proposed method shows high performance on the basis of classification accuracy and efficient execution while comparing with two other state-of-the-art methods.

Texture Measurement Through Local Pattern Quantization for SAR Image Classification

A novel local pattern based classification algorithm for SAR image is proposed in this paper. The proposed method initially quantizes homogeneous and non-homogeneous patterns within the moving window. An operator is constructed to quantize local patterns. Quantized patterns are then used for measuring texture around the central pixel within the moving window. The ISODATA algorithm is used to classify texture transformed image. The proposed classification method is robust to speckle noise, computationally simple and does not need to set any predefined parameter for classification. The validation of the method is done on RISAT-1 and RISAT-2 data.

Satellite-Based Estimation of Instantaneous Radiative Fluxes Over Continental USA – a Case Study

Instantaneous radiative flux components across different climatic regions of USA were computed using Moderate Resolution Imaging Spectroradiometer (MODIS) land, atmosphere and geo-location products for clear sky pixels, colocated in time and space with flux towers. Satellite derived fluxes were validated using in-situ SURFRAD flux tower data of the respective sites. Significant positive correlation was observed between simulated and observed fluxes especially for incoming shortwave (r = 0.94), incoming longwave (r = 0.91) and outgoing shortwave (r = 0.89) radiation. The root mean square errors (RMSE) were 50.7, 10.41, 18.54 and 33.50 Wm−2 for incoming shortwave, outgoing shortwave, incoming longwave and outgoing longwave fluxes respectively with their corresponding relative RMSE of 0.11, 0.10, 0.06 and 0.08. The D-index and modeling efficiency (ME) varied from 0.87 to 0.96 and 0.61 to 0.85 respectively, indicated good performance of the present method. The coefficient of residual mass (CRM) for all the fluxes yielded values close to zero except incoming longwave radiation. Based on the statistical analysis and accumulative score highest rank was obtained for incoming longwave flux followed by outgoing shortwave, incoming shortwave and outgoing longwave fluxe. From the rank values it can be concluded that the present model predicted all the fluxes satisfactorily except the outgoing longwave flux.

Near-Real-Time Availability of Ocean Heat Content Over the North Indian Ocean

Ocean heat content (OHC) is an important parameter in determining the heat flux in the ocean-atmosphere system, which can influence weather systems such as cyclones and monsoons. Hence, regular monitoring of OHC is required, which needs continuous subsurface temperature profiles. Due to the scarcity of in situ temperature profiles in space and time, remotely sensed sea surface temperature (SST) and sea surface height anomalies (SSHAs) are employed in the computation of OHC in the Indian Ocean. OHC derived from in situ temperature profiles from ARGO floats along with collocated SST, SSHA and OHC climatology during the period 2002-2012 are used to estimate OHC700 (heat content up to 700-m depth), using an artificial neural network model. The estimated OHC700 is validated and is found to be significantly correlated with the observed OHC700. Using this approach, OHC700 is being estimated daily on a near-real-time basis, and the products are available at http://bhuvan.nrsc.gov.in/data/download/index.php.

Cartosat-1 Image Segmentation Technique for Shade Tree Crown Density in Tea Gardens of East India in Relation to Terrain Geometry

One of the factors determining tea quality is shadow casting by the shade trees. Besides regulating incoming solar radiation shade trees also helps maintaining the moisture in soil and nutrient recycling. However the optimum shade density depends upon the elevation, slope and aspect. In the present study image segmentation technique was employed on Cartosat-1 data to capture the vertical crown density of the shade trees. Significant positive correlations (r 2 =0.91) were found between observed and measured vertical crown density. Based upon the crown density the tea gardens were classified. Further the relation between crown density and terrain parameters has been analysed. Significant negative correlation was observed with elevation (-0.590) and slope (-0.627) which indicates that to increase in elevation and/or percent slope the shade density decreases.

Application and comparison of advanced supervised classifiers in extraction of water bodies from remote sensing images

Abstract Water body extraction plays an important role in monitoring and assessing the existing water resources. It is a complex process that may be affected by many factors. This paper examines the major and advanced supervised classification approaches and ventures into the effectiveness of these techniques in extraction of water bodies from satellite images. The different classification techniques used for this purpose include support vector machine, artificial neural network, K-nearest neighbor, discriminant analysis and random forest. Commonly used normalized difference water index technique has also been examined in the study. Comparisons have been drawn among various variants of these methods and the accuracy in each case has been recorded. Each classification technique has been applied on input images from three different satellite sensors of varying spatial and spectral resolution, to compare their performance on different data sets of three different study areas. The study has found that supervised classifier can extract water bodies with a good accuracy from remotely sensed images even with a fewer number of labeled samples. Additionally, it is seen that the linear classifiers also yield good accuracy in extracting water bodies across different sensor’s data.

Standalone Open-Source GIS-Based Tools for Land and Water Resource Development Plan Generation

Land and water which are extremely important natural resources are depleting at a fast rate. This has triggered the need to conserve these resources and utilise them optimally. Proper planning that involves strategies for achieving a desired set of goals, and management of these resources are required for sustainable development of any geographical area at watershed scale. Remote sensing and Geographic Information System (GIS)-based techniques have the potential to generate as well as to analyse geospatial data that serve as key inputs for generation of land and water resource development plans at the watershed level. However, planning authorities are adopting conventional methods for planning due to lack of GIS knowledge which is time consuming. Thus, the proposed customised open-source GIS tools not only help in bridging the knowledge gap but also help in generation of land and water resource development plans in short time. MapWindowGIS is a unique standalone open-source GIS component that can be customised through dot net programming. In the present study, MapWinGIS is used to develop Land Resource Development (LRD) and Water Resource Development (WRD) plan generation tools. These tools employ a set of logical conditions over a set of input layers to produce WRD and LRD action plans for a chosen area. A spatial database, pertaining to the Kuchai block of Saraikela district, Jharkhand India is used for generation of land and water resources development plan. These plans involve the generation of alternate land use and demarcation of areas suitable for artificial recharge. Thus, the tool enables to integrate together spatial data from diversified sources in order to analyse and produce meaningful information for decision makers to support their planning activity.