Home  EnviroNews  International Conferences  Picture Gallery  Sponsor  Contact  Search  Site Map

Vol. 18 No. 4 - October 2012

Remote Sensing and GIS for Forest Monitoring and Management

By: S.P.S. Kushwaha


The problems involved in maintaining a sustained supply of forest resources for the present day needs and future demands of the mankind have made the forest managers conscious about the compelling need for rational utilization of these resources. The conventional methods of forest resources assessment and monitoring are time and cost-intensive. Many a time they do not match with the resource dynamism and hence, become obsolete by the time the results are available. The advent of remote sensing and geographic information system (GIS), and global positioning system (GPS) technologies has revolutionized the forest resources assessment, monitoring, management and has reduced the time and cost considerably. Remote sensing is the science and art of obtaining information about an object, area or phenomenon through analysis of data acquired by the device that is not in contact with the object, area or phenomenon under investigation (Lillesand and Kiefer, 2006). It is also said to be the practice of deriving information about the earth’s land and water surfaces using images acquired from an overhead perspective, using electromagnetic radiation in one or more regions of the electromagnetic spectrum, reflected or emitted from the earth’s surface. Some of the highlights of the remote sensing technology are stable sensing platforms, synoptic coverage, high frequency of observations, and real-time images available in spatial form and on multiple scales. It is perhaps the only technology, which allows retrospective evaluation of forest resources.


Remote sensing, as such a multidisciplinary discipline, is utilized in the inventory and monitoring of the forest resources. All remote sensing systems including aerial cameras capture radiation (signature) in different wavelengths reflected or emitted by the land and water surface features and record it either directly on the film as in case of air photos or on a digital medium like tape, and other digital media. As no two objects in nature are theoretically ditto, their signatures are unique. This property of the objects is exploited in remote sensing to differentiate the objects from one another. The reflectance from vegetation is controlled by leaf pigments, cell structure, and the leaf water content. The radiation absorbed in red region is primarily used for photosynthesis. In healthy vegetation, both absorption and reflectance are more pronounced. Diseased and senescent vegetation shows lesser absorption as well as reflection in red and near-infrared regions and a higher overall reflectance in blue and green regions. These spectral properties of vegetation are exploited to detect their type and condition through image interpretation. Forests are one of the most conspicuous terrestrial features on the planet earth. Identification and mapping of forests using remote sensing techniques is thus, relatively easy. Satellite remote sensing is extremely popular in forest surveys.


Satellite Remote Sensing

During the World War-II, the use of electromagnetic spectrum was extended from visible and infrared to microwave regions and this is considered as a major milestone in the history of remote sensing. The beginning of space-based remote sensing dates back to 1891, when Germans developed rocket-propelled camera systems. But it was in 1957, when Sputnik-1 took the first photograph of the earth from satellite. Systematic earth observation from space started with the launch of Explorer-1 in 1959 and meteorological satellite, TIROS-1 in 1960. The launch of Earth Resources Technology Satellite (ERTS) in 1972 dawned a new era in remote sensing. This was the first satellite available for systematic and repetitive observations of earth’s land resources. Landsat-1, 2 and 3 carried multispectral sensors operating in 0.5-1.1µ wavelength range and had 79m x 57m spatial resolution. The satellite data provided by these satellites facilitated the identification and mapping of broad forest types and canopy densities. Landsat-4 and 5 carried both MSS and Thematic Mapper (TM) sensors and provided low as well as medium resolution (30m x 30m pixel) imagery. The Landsat TM imagery changed the foresters’ outlook about the forests. The availability of 20m and 10m resolution imagery from French satellite, SPOT later significantly advanced remote sensing applications and brought image interpretation close to virtual reality.


The Indian Space Research Organization (ISRO) launched a number of Indian Remote Sensing Satellites (IRS) with capabilities similar to contemporary earth observation satellites. The launch of Indian Remote Sensing Satellite (IRS-1A) in March 1988 marked a new era in the history of satellite remote sensing programme in India. Subsequently, the IRS-1B, IRS-P2, IRS-P3, IRS-P4, IRS-1C, IRS-1D, TES and IRS P6/Resourcesat-1, and Resourcesat-2 were launched. Both IRS-1A and 1B carried 72.5m and 36.25m spatial resolution sensors on-board and provided not only the continuity of satellite data from Landsat Programme (of U.S.) to indigenous one but also an opportunity for Indian scientific community to test their data for forest resources inventory and monitoring. While the earlier two sensors are meant to facilitate in locale-specific intensive resource inventories, the WiFS (Wide-Field Sensor) and AWiFS (Advanced Wide-Field Sensor) data facilitated the assessment of the land and water features and phenomena encompassing large areas. The Indian Remote Sensing Programme has come a long way and is all set to grow further through more advanced sensors with improved capabilities of information generation.


The availability of satellite data in digital form provides a whole lot of flexibility with respect to its use. Single band black and white as well as false colour composite (FCC) images can be used for interpretation. Since single imagery covers large area, the broad forest features encompassing larger areas could be studied. The use of multi-date imagery provides information on phenological conditions of the forests i.e., evergreen or deciduous as well as on the extent and the distribution of forests with the passage of time. Over time, the visual interpretation has given way to digital interpretation. It has been found that a combination of supervised and unsupervised techniques i.e. hybrid methods yield better results. In India the first attempt to categorize forest cover types by digital classification of satellite data was made in 1978 for Nagaland delineating temperate evergreen, tropical evergreen, tropical semi-evergreen, tropical deciduous, bamboo, degraded forests, shifting cultivation and permanent cultivation. Many studies on forest cover mapping have been done in India so far using satellite imagery. NRSA carried out the first-ever forest cover monitoring using 1972-75 and 1980-82 timeframe satellite imagery, showing very high rate of deforestation (NRSA, 1983). In general, forest cover mapping has been attempted more frequently than type mapping. Selecting a time of the year when maximum differences occur due to phenological changes such as leaf fall, flowering etc. improves the capability of satellite data in forest type delineation.


Microwave Remote Sensing

The use of microwave imagery in forestry/vegetation sciences is mainly driven by the fact that microwave remote sensing is capable of providing data at any time of the day/night and more than that is the capability of microwaves to penetrate the atmosphere under virtually all conditions. Depending on the wavelengths involved, microwave energy can see through haze, light rain, fog, snow, clouds and smoke. Hence, microwave remote sensing has some edge over optical remote sensing. Moist vegetation returns more signal than dry vegetation. Also like-polarized (HH or VV) sensing penetrates vegetation more than cross-polarized (HV or VH) microwave remote sensing. Likewise more energy is returned from crops having their rows aligned in the azimuth direction than from those aligned in the range direction of radar. Radar imagery has been widely used in qualitative and quantitative forest stratification. The backscattering from forested areas has been found to be dominated by tree crowns consisting of foliage and branches. The K- and C-band have been found to be sensitive to low biomass level while P-band is sensitive to high biomass forests. The backscattering from broadleaved forest stands is normally more than that from coniferous forests. Many studies have related various forest stand parameters like tree height, density, age, timber volume, biomass etc., with radar backscatter with varied degrees of success.


LiDAR Remote Sensing

LiDAR (Light Detection And Ranging) remote sensing is a breakthrough technology for forest resources inventory. It offers a great potential for forest conservation and management. The advantage of using LiDAR is that it provides three-dimensional data. If cautiously planned, LiDAR can form one of the most scientific and accurate means of forest management. The various advantages of LiDAR technology are higher accuracy, weather independence, capability of canopy penetration, lesser time needed for data acquisition and processing, minimum user interference. Besides, laser-derived images help in terrain visualization. As vertical component (z-axis) measurement is the backbone of LiDAR technology, this characteristic is exploited in a very straight forward way for tree height estimation. Tree canopy height is obtained by subtracting the elevations of the first and last returns. Vegetation height when coupled with species composition and site quality information, serves as an estimate of stand age or successional stage. Like simple height estimate, the vertical distribution of laser returns provides basis to classify vegetation, and to estimate other important canopy characteristics such as canopy cover, crown volume (foliage, trunk, twigs, branches etc.). LiDAR data provides input for estimation of aboveground biomass with high accuracy. The combination of LiDAR and satellite remote sensing data could be very useful for describing biodiversity and monitoring changes in biodiversity.


Forest Type and Canopy Density Mapping

The utility of the remote sensing data for forest canopy density mapping and monitoring on various scales is well established by now. The nationwide forest studies carried out by NRSA and FSI have amply demonstrated the scope of satellite remote sensing in forest mapping and monitoring. Multispectral data such as that from IRS LISS-4 with 5.8m, IKONOS with 1m, and QuickBird with 0.61m spatial resolutions provide an unprecedented opportunity to monitor the forests. Barring single species dominated forests and forest plantations, majority of the Indian forests are highly heterogeneous. This makes their differentiation, delineation and mapping a difficult task. Problem gets further compounded in case of forests that are located in hilly and mountainous regions on account of topographic effects. Forest type mapping from satellite imagery has been attempted in past with varied degree of success. Local level studies backed by intensive ground truth have generally resulted in more number of forest type categories. Except for the small scale map prepared by Champion and Seth in 1968, showing sixteen major forest groups, India did not have a forest type map until recently. This year, the Forest Survey of India, for the first time, has come out with a satellite image-based forest map with 180 types.


Wildlife Habitat Evaluation

Remote sensing can be applied to wildlife habitat inventory, evaluation and wildlife census. Remote sensing not only provides spatial data but also allows us to compare temporal variations in the spatial data, essential for wildlife management. While ground survey methods such as counting animals, trapping, collection of droppings, investigations of feeding sites as well as ground mapping of habitats will always be useful, remote sensing can supplement or considerably replace tedious ground surveys. Ground methods have limitations as whole area can not be accessed in one go in many of the cases and the information collected may not be as accurate as is possible through remote sensing. The GIS-based modeling of species-habitat relationships is one of the popular forms of habitat suitability analysis (Singh and Kushwaha, 2011). The GIS output is a map depicting habitat suitability for any wild animal. The map can guide decisions regarding habitat preservation priorities, forest/land management practices, or sites for reintroduction of endangered species. As more and more data and information is being collected day by day by wildlife managers, world over, a need is being felt to develop a Wildlife Information System (WILIS), integrating wildlife species databases in spatial/non-spatial formats, habitat suitability models and rules/guidelines for habitat evaluations. A WILIS could be operational at national level with linkages to regional and local databases. One particular study done in Chilla Sanctuary showed US$ 100 as the cost of per square kilometer habitat evaluation.


Timber Volume/Growing Stock Inventory

Timber volume and the total growing stock are the key information required for the forest planning and management. Remote sensing data facilitates in the stratification of forests, which in-turn reduces the sampling error and allows the growing stock assessment with fewer samples. Most of the territorial forest divisions have working/management plans and these are revised every ten years. Conventional methods have limitations and hence, space technology should be adopted to achieve objectives envisaged in the National Forest Policy. The satellite image-based forest stratification can be correlated to the actual on-ground timber volume/growing stock or biomass using two-stage inventory design (Köhl and Kushwaha, 1994). For large areas, it is advisable to go for multi-phase sampling techniques. In a multi-phase design, visually or digitally classified imagery makes the first stage followed by large-scale image interpretation and ground measurements. The multi-phase sampling design reduces the cost considerably. Overtime, high resolution satellite imagery has significantly replaced aerial photos in forest in growing stock and biomass inventory.


Plant Richness Assessment

India, with a geographical area of 2.4 per cent of the world, has about 8 per cent the world’s total biodiversity. The country is very rich in biodiversity with 45,000 plant and 75,000 animal species. Hence, it is called as a mega-diversity region. The threat to biodiversity in India is due to the over-exploitation of plant and animal resources as well as their habitats resulting in the fragmentation of habitat and creation of considerable landscape heterogeneity (Singh and Kushwaha, 2008). Remote sensing and GIS can contribute immensely in biodiversity assessment at regional to global scales. Remote sensing can play a very useful role in assessment of bio-rich areas, which could then be related to actual values of biodiversity, got through ground methods. This can be done through landscape characterization wherein a digitally or visually classified image is taken as direct input (Behera et al., 2005). Many landscape parameters viz., porosity, patch size and shape, interspersion and juxtaposition etc. have been said to have a direct relationship with a variety of vegetation features like biodiversity, physiognomy, composition and other stand parameters. Geospatial analysis of a forest landscape can take care of many factors that persist within each ecosystem. Vegetation type maps are prime inputs for biodiversity assessment at landscape level. The impact of scale on biodiversity assessment is significant; higher scale results in better assessments (Kushwaha et al, 2005).


Forest Fire Risk Assessment

More than ninety five percent of the forest fires in India are man-made. Forest understorey is burnt year after year by local people for deriving benefits like cattle grazing, fodder, accessibility for timber/firewood extraction etc. Recurrent burning hampers forest regeneration and arrests plant succession. About 65 percent of the India’s 64 million ha forests are broadleaved deciduous and ca. 5 percent are coniferous; thus, about 70 percent deciduous forests happen to be fire-prone. The estimated average annual tangible loss due to forest fires in the country is of the order of approx. US$ 100 million. The Himalayan coniferous forest comprising of fir, spruce, cedar, chirpine and blue pine etc. are highly prone to fire. The factors governing fire occurrence and spread are: fuel loading (type and moisture), temperature, humidity, wind (both speed and direction), slope, aspect and the accessibility. A rapid assessment of the forests for their proneness to fire could be worked out using a combination of information generated from remote sensing and other means. The information on wind is generally seldom available because of the poor network of meteorological stations in India. A dynamic fire spread model using a combination of biophysical parameters including wind direction, speed and live fire events could be worked out to predict the fire spread.


Forest Litigation

Remote sensing and GIS applications have assisted the judiciary for quite some time now to decide on the disputed cases of land ownership in the country. In one such case between Sanjay Gandhi National Park authority, Mumbai and the encroachers, wherein considerable area along the park periphery had been encroached by local people over past several decades, the High Court at Mumbai, based on the evidence produced by remote sensing gave the verdict in favour of the Sanjay Gandhi National Park and directed Government of Maharashtra to evict the encroachers. In another similar case in Maharashtra involving farmers of Dhule district and the forest department, wherein farmers had encroached during past three decades sizeable reserved forest area and were practicing agriculture, the remote sensing and GIS-based evidence showed that the disputed area actually was well within the reserved forest boundary and hence, the High Court in this case too decided in favour of the Forest department. In yet another similar case, where fisher folks had encroached a part of the Jambu island in Sunderban Biosphere Reserve for fish drying, the Supreme Court of India gave the verdict in favour of the Biosphere authorities after untiring efforts of the Director, Sunderbans Biosphere Reserve by producing concrete evidence through time-series analysis of satellite images.


National Forest Information System (NAFIS)

Responsibility for 64 million ha of dense, open forests and forest plantations across the country falls under a great many jurisdictions and management agencies including government organizations and the individuals. Forestry information is gathered in different ways, for different uses and is stored in different formats and locations. As a result, accessing and integrating this information is extremely complex. There is need for a national forest information system (NAFIS) for effective information and communication management. NAFIS is envisaged as an internet information gateway to forest information resources from across the country. The NAFIS should address these significant and wide-ranging differences by adopting international standards and by building a distributed network of servers and applications which allow access to forestry information held by independent agencies. The NAFIS should provide web tools ranging from simple portrayal to sophisticated analyses, to users from anywhere in India or abroad, which means that users can discover, interact, integrate and display the authentic and accurate information on India’s forests and forestry. NAFIS should promote standards for enhancing the interoperability of multiple information resources.


Future Prospects

The need for higher spatial, spectral and radiometric resolutions for forest types and species (associations) has been emphasized over time. Gradually, the mapping and monitoring scenario is expected to be a lot better than ever before. The spatial, spectral, and the radiometric resolutions have already improved considerably. Currently many of the sensors provide spectral resolutions of the order of about 10 nanometers per band. Sensing using continuous spectra is expected to help not only in better species identification, association/formation and forest/vegetation type level mapping but also result in higher accuracy in timber volume and biomass estimations by highlighting the subtle differences in the physiognomy of the vegetation. The hyperspectral imagery is providing insight into the state of biodiversity and the vegetation continuum across ecosystems and the landscapes in an unprecedented manner. Ground penetration radars are already helping the scientific community in belowground biomass Assessment.


Forestry and Ecology Department, Indian Institute of Remote Sensing, Indian Space Research Organisation, Dehradun 248001, Uttarakhand, India. <spskushwaha@gmail.com>


This article has been reproduced from the archives of EnviroNews - Newsletter of ISEB India.

Home | EnviroNews | International Conferences | Picture Gallery | Sponsor | Join/Contact | What others say | Search | Site Map

Please report broken links and errors on page/website to webmaster@isebindia.com