Role of Remote Sensing in forest management

C. B. S. Dutt
Head, Forestry & Ecology Division,

V. Udayalakshmt
Scientist/Biometrician, Forestry & Ecology Division,
National Remote Sensing Agency, Balangar, Hyderbad. India

A. S. Sadhasivaiah
Chief Conservator of Forests (WG),
Karnataka State Forest Department, Bangalore.

Abstract

The satellite aplications for effective forest management on a more scientific basis commensurating with the priorities set at state, District and Micro levels studies. The shift in priority of forest management towards ecologically sustainable forest resources management call for reliable spatial database with a provision to update and retrieve for management decisions at various levels. The application of satellite data for various priorities & objectives leading to resources assessment have been discussed. The utilisation of GIS for data base creation and requirement of forest resources information system involving effective inventory data analysis packages supporting volume yield and cull factor analysis has been discussed in detail. The concept of NRIS in the country using remote sensing has been emphasised.

Introduction

Historically Forestry has been concerned mainly with the assessment of timber resources and the management and utilisation of closed forests for the production of wood. Actention was occassionally given to the other resources of the land sustaining closed forest. In the 19th century some working plant in continental Europe considered not only timber production but also wild life and or for protection. Forestry in
urope based on conifers was also aidly expanded, but in the tropics, subtropics and north America, the closed natural forests were increasingly exploited for timber.

During the british regime, the forest survey and sound forest management was extended to all lands bearing trees in India and to minor forest products (e.g. gum arabic). As early as in 1856 a forest department was established in Burma by the British and the concepts of environmental Forestrywere being applied in India before the end of the
nineteennth century. As viewed by the FAO established by the Bretton wood conference in 1944, forest have become widely recognised as ” all lands bearing a vegetation association dominated by trees of any size, exploited or not capable of producing wood or other products, of exerting an influence on climate or on the water regime or providing shelter for live stock and wildlife (Loetsch and Holler 1964). It was not, however until the oil crisis of 1973 that world wide interest was re-focussed on the long-standing importance of forests as the major source of energy in may countries. Arecent European study (ECE, 1986)

Suggest that over 40% of one votoe european Forests is used as a source of energy and that energy remains the single most important use of wood in volume terms.

With the uncertainity of energy prices in the future and of the growth world wide in the consumption of energy, predicted to exceed 2% a year , the demand for wood energy can be expected continued increasing. Within a decade the most urgent need of many local communities in the developing countries will be the massive harnessing of their resources to renew forest supplied through plantations and agro- Forestry, and where preactised the “Slash – and burn” of natural forests based on short cutting cycles. In response to the situation remote sensing can be expected to be used increasingly to collect urgently needed data, especially as related to monitoring changes in forest cover, assessing landus and frorest land degradation, evaluationg the productivity of the land and providing
information not only for forest inventory but also for direct inputs to forest management and strategic planning.

The increased awareness due to mounting population exerting pressure on forests for fuel and timber besides grazing for cattle. This has led to most of India’s forest left with poor carrying capacity. The forests began to dwindle. And now the country have much less of its land area under forest cover than is required to maintain its environmental
stability and ecological security. The extent of biotic pressure on forests could be judged from the fact that with less
than 2% of the total forests areas in the world, India supports 15% of the world population and nearly 14% of the cattle population.

Considering the fact, India is currently carrying out biennial monitoring of forest cover using satellite data on 1:250,000 scale. However the system of forest management in India is almost 120 years old and remained as a state subject. The policy formulation and strategic planning is at the central level. For effective
management of forest resources the various basic requirements and priorities are essentially of three tier system i.e., State level District level and Micro level information requirements. The priorities are (1) distribution (3) plantation monitoring (4) estimation of forest growing stock (5) forest inventory and volume estimation (6) preparation of stock tables and yield calculation (7) preparation of treatment/ zonation areas (8) assessment of ecological and bio-diversity (9) forest change and conversion studies (10) Forest damage assessment due to forest fires (11) Grassland identification & productivity assessments (12) biomass and fuel wood assessments (13) GIS applications (14) implementation of forest resources in formation system and (15) future thrust areas on forecasting and prediction models using remote sensing and other collateral data.

The present scenario of remote sensing applications in forest management is widely applied in the areas of:

Forest Cover Monitoring/ Surveillance

The forest cover at the national level is being biennially monitored using remote sensing data and it is estimated that India has 19.47% of forest cover (1989-91) out of the total expected 33% of forest cover as per India’s forest policy. So far Forest Survey of India (FSI) has broughtout sucessfully four assessments on
the status of forest cover in India (Table.)

This information is more relevant in the context of assessing rate of degradation grossly at the state/district level in terms of closed and open forests. However, the extended use of this information for effective
utilisation for the management purpose at the state level could be better achieved by supplementing the data each reserve forest boundaries thereby enlisting the districts having ore degraded areas. This approach enables to prepare treatment area maps at the state level which would serve as a strategic plan inputs for prioritisation and categorisation of areas for better silvicultural practics.

Table 1
Status of Forest Cover in India
(Based on Satellite Remote Sensing)
Area in million Hectares within brackets

The extended use surveillance maps for the treatment area maps would need a serious consideration by all the state forest departments for prioritisation:

CLOSED FORESTS (40% and above crown density) treatment area I for consideration to conservation zone.

OPEN FORESTS (less than 40% crown density) treatment area II for consideration as forest produce zone for careful management.

DEGRADED FORESTS (less than 10% crown density) treatmen area III for gap planting, JFM activities by peoples involvement.

Thus for the above treatment area maps the satellite based remote sensing especiallyusing the digital methods would prove to be an effecctive tool and in generation of information within a short span of ltime and a digital data base.

Forest Type Mapping and Assessment of Distribution:

India has been divided into broadly sixteen forest types by Champion and Seth (1968) based on rainfall and altitude. However, with changing climate and man’s impact on environment particularly in forests, change the composition of tree species resulting spatial disturbances on the occurrence of forest types. In this context the study of spatial ditribution of forest types. In this context the study of spatial distribution of forest types would help greatly in forest management for accounting the changes that have resulted in the chaning quality and composition of forests. The case studies carried out in parts of Western Gahts Viz. Coorg district and Uttar Kanara District, brings out the feasibility of satellite aplications in preparation of spatial forest maps at the block level. The changing scenario of natural forests into artificial forests could well be discerned by generation of a baseline data of forest types which are independent of each district. The availability of such maps have greater ecological significance and provide better insights to the working plan officials as to what type of species need to be introduced by replacement of natural forests or regener4ation of degraded forests in the area on a more scientific and reliable distribution at the block level on 1:50 ,000 scale would prove an essential component in the scientific management of forests.

Forest Stock Mappping and Preparation of Working Plan Inputs

The satellite data analysis is found adequate for differentiating three levels of stocking i.e. closed (40 and above), open (10-40%) and degraded (below 10%) crown cover classes on 1:250,000 scale. However, the property of stocking levels generally required at the working plan level is of 20% crown cover intervals, The adoption of spectral based stock maps appear get saturated beyond40% crown cover, However, in certain areas of Madhya pradesh it is found, based on tone and texture associated with spectral separability as a function of time has been found suitable to provide 40 60% stocking level in addition. This study indicated greater scope for future exploitation and utilisation of various algorithms for optimising time window and spectral bands may be feasible to explore satellite based forest stock map preparation. The studies using NDVI as a function of stocking level needs review as it provide information pixel level on the vigour of the canopy with lwss degree structural relationship of canopy closure. The studies within NRSA indicate the saturation of NDVI levels beyond 40% cover is observed. However the development of suitable algorithms using textureal function might prove to be an effective method for quick assessment of stocking levels. The terrain contributed noise due to change in aspect and elevation could also be integrated by development of suitable digital elevation models as background to minimise the noise for preparation of dependable stock maps. While the spectral based models appear feasible the spatial resolution would contribute only interms of small patches and their associated stocking levels. With the experience on IRS LISS II data, is adequate for accounting canopy closure level as that much spatial resolution would be minimum required as a spectral function from a representable area of the forests for coorelations on the ground cover conditions.

The studies carried out in Uttar Kannara by comparitive evaluation of detailed stock map prepared from aerial photograbphy and confirmity and confidence in adopting satellite data for preparation of stock maps with suitable modifications. However, the satellite based stock maps showed. The aggregation of patches due to the limitation of resolution and contribution of terrain noise, In any case the level of information obtained with respect to stocking is of the order of 70% and above. The study broughtout the scope of reconnaisance level stock map preparations at the district level for preparation of management plans.

Forest Inventory and Sampling

The forest enumeration and surveying is generally carried out at the district level for detailed working plan inputs preparation generally on 1:15,840 (4′ = 1 mile) on predecided percent area enumeration (5% , 10% etc.) The working plan enumeration inputs is based on chain surveys across the forests on a systematic way. The conventional way of systematic sampling suffer due to accounting various strata and burdened with intense field work & expensive procedures, The satellite data facilitate in generation of primary stratifucation units at the district level either throudh use of forest density maps or forest type classification it is possible to stratify the area into homogenous forest strata. Based on the apriori knowledge the minimum sample size can be selected and the field inventory can be accomplished by suitable proportional allocation of samples in all the strata The stratified sampling procedures provide a reliable quantification of forest resources depending on the objective set during the inventory.

The study carried out in parts of western Ghats of Karnataka demonstrated the feasibility and to build volume from the enumerated data on a time & cost effective manner.