Home Articles Role of Remote Sensing in forest management

Role of Remote Sensing in forest management

ACRS 1994
Forestry
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

Forest Category 1972-75 1981-83 1985-87 1987-89 1989-91
Dense/ 14.12 10.99 11.51 11.71 11.72
Closed (46.45) (36.14) (37.84) (38.50) (38.55)
Open 7.38 8.41 7.83 7.60 7.61
(24.28) (27.65) (25.74) (24.99) (25.04)
Mangrove 0.10 0.12 0.13 0.13 0.13
(0.30) (0.40) (0.42) (0.42) (0.42)
Total% 21.60
(71.03)
19.52
(64.20)
19.47
(64.01)
19.44
(63.92)
19.47
(64.01)

 

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.

ACRS 1994

 

Role of Remote Sensing in Forest Management


The Forest Volume Estimation and Generation of Volume Tables

The satellite based multi phase approach forest inventories with defined objectives provide ample data for further processing and computation of volume and yield tables. The enumerated data during the inventory could be systematically organised strata wise ad a suitable local and stand regressions could be generated fr further computation of species wise volumes. The predominant species volume equations thus generated would form the base for computation of tatal standing volume based on the plot enumeration data. The inventory data analysis in the Uttar kanara Circle using satellite base stratification sampling approaches have yielded better observations in development of local and stand tables and to compute stand volume. The standing volume information through inventory data analysis would form as a baseling data to bring out correlation with ground crown density maps. This in a way to say once the relationships are established which are generally local specific with respect to volume and density it is possible to estimate the total growing stock of the area by generation of stack maps and conversion through establised volume functions. The experience in generation of such stock to volume estimations in western Gats have shown promissing to explore to avoid cumber some, field inventories. The study also discussed in detail the importance of cull factor for estimating merchantable yield.

Stock Tables Prepartion and Yield Calculation

From the stratifield sampling and inventory data analysis it is feasible to calculate and generate stock tables with respect to number of species diameter and height wise distribution. Such an information with the available knowledge of minimum girth area for extraction it is possible to calculate the yield and also to compute the annual coupes based on equiproductive areas. The technique of data analysis and generation of stock tables developed in parts working plans project have revealed greater scope for extending towards management information system specifically for the working plans at district level on stock tables and yield calculations.

Ecological Consideration and Zonation of Forests

With the increased pressure on forest all over the world as well as in India the growing concern of forest management has shown shift in its priority from production foresty to conservation Forestry. In the light of thi the scientific management of forest and categorisation of the area into different zones have been adopted by the Karantaka State forest department as part of Western Ghats Eco-restoration project. Towards this NRSA and Karnataka Forest Department jointly carrying out a project in a area of 10,280 sq. km in Uttar Kanara forest circle by generation of multi-thematic information on 1:25,000 scale using aero-space techniques. The thematic information generated are forest density and height maps, forest type maps, slope and aspect maps, drainage map upto first order channels, volume class maps besides consultation of contour maps with the available topographical sheets. The availability of such an enormous data base when feed into the Geographic Information System enhances interpretability of the data and to classify the area into different zones required for effective forests management practices on a scientific footing. Accordingly the forests are proposed for categorization of zones for specific management practices, the zonation is the process of describing the physical consequences of the management plans, which have been understood for sustainable forests resources in the area. It is essentially link between management practices. The zonation in the process of describing the physical consequences of the management plans which have been understood for sustainable forests resources in the area. It is essentially link between management objectives and physical operations on the ground.

Accordingly using multi-thematic information the zonation of any area could be prepared with the below given logic. (Table 2)

The below given process com The study carried out in the parts of Western Ghats showed adequacy of the tool and approach.


Table 2 Forest Zonation Matrix

RF D H(M) Slope% Elevation (M) Village Type
Zone 1 + 4.5 >20 >50 High Nil+ Managroves Grassland
Zone 2 + 3.5 >20 >35-50 Medium Sparse/Nill Any Type
Zone 3 + 2.3 >20 >35-50 Medium Inhabited Any Type
Zone 4 + <0.25 All >35 Rolling Habitated -Do-
Zone 5 Reve All Lands Falling in “C” and “D” Lands W. L

 

Remote Sensing and Biodiversity Studies

The forest management while through its forest inventory and zonation of the area can revolve ecologically sensitive and diversity wish rich areas through use of satellite data applications in conjunctive use of other ancillary data discussed as above in the zonation. It is estimated that the tree diversity is estimated to around 200 species in the northern parts as Western Ghats exhibited a great diversity between the plots and along the gradients. The extended application of forest inventory and enumeration including shrub layer and herb layer would enhance forest management system for preservation of ecologically rich zones and to account diversity of the areas. The initial probabilistic diversity zones should be narrow down using satellite data and the consistency of patchiness as a function of landscape dynamics could be used as an element for bio-diversity studies.

Forestry Conversion Studies

With the rapid destruction of forests and encroachment the use of multi-temporal satellite data using digit analysis procedure provide spatial change maps by image differencing methods and logical operations. The study carried out using multi-temporal satellite data of 1983 & 1993 in parts of Andhra Pradesh of Adilabad District showed the area decreased upto 25% of the total area studied whereas improvement in the quality of the area is only 6-7% out of the 1,512 sq. km area studied, The methodology recommends for detailed investigations at district level for accounting forest changes over the years within the reserved forests areas.

Forest Fire Damange

The use of multi-mission IRS data has demonstrated amply the feasibility to identify forest ground fire damaged areas with the combined use of IRS 1A and IRS 1B. The studies carried out during 1991 in parts of Nagarghole wild life sanctuary facilitated accounting damage of an area of 68 sq. km. Quite recently in the month of March 1994 the central part of the India in the Simplipal reserve forests the extent of fire damage was assessed using IRS 1B and estimated the ground fire damage to an area of 600 sq. km against 200 sq. km reported. The use of in season data would enhance capability in identifying hot spot areas which are annually prone to fires could be of use to the forest management for mounting fire operation measures during the vulnerable periods.

GIS Applications in Forestry

The scientific and effective forest management information could be operationalised using the GIS approach at district level as a functional unit. The organization of data base, data inputting, updating, retrieval and analysis for specific purposes and to obtain outputs apparently is feasible cost effectively at the district level rather than aiming at state level as a central facility. The data base in GIS should necessarily preceded by generation of reliable spatial maps preferably on 1:25.000 scale to begin with which can be later inputted to GIS for proper data utilization and to obtain outputs with greater degree of accuracy relevant for district level management decision. The organization of data feasibility adopting micro scale to macro scale may be a viable approach rather than adopting top to bottom approach stasr6ting with coarse resolution data which may fail to reveal any meaningful outputs required at the working level. The organization and aggregation of data base at district level would enhance efficient forest working most needed in the district level. The use of GIS at taluk level for preparation of zonation map revealed micro scale details without sacrificing accuracies for management decisions.

While the GIS data base is being parallel evolved simultaneously with the improved spatial maps the requirement on development of forest resources information system at the district level in the form of MIS should form a major management aspect which facilitate aggregation at the state level for strategic decisions The concept of National Natural Resources Information System (NRIS) as part of NNRMS activity of Dept. Of Space would be the fore runner in evolving a reliable operational forest management information system.

References

  • Loetsch, F and Haller, K.E. (1964) Forest Inventory, BLV Verlagsesselschagt, Munich.
  • ECE (1986) The ECE Timber Committee year book -XL (1) United Nations Economic Commissio for Europe.
  • MOENF (1993) The state of Forest Report 1993 GOI.
  • Champion, M.G. and Seth, S/K. (1968) Revised Survey of Forest types of India , Govt. of India publication.
  • Gadgil, M., Sinha M. Pillai J. (1989) A Biomass budget. Final Report of the study group on Fuelwood and Fodder, Planning Commission Govt. of India.
  • Unni, N.V.;M. (1990) Space and Forest Management. Procedings of space and forest managemt, Special Current Even Session 41st, IAF Congress, Dresden, Germany.
  • Weigand , C.L and Richardson A/J. (1990) Use of spectral vegetation Indices to inter leaf area, ET and yield : 1 Rationale Agron J. 82: 623 – 629- Annexure -1
  • Hush, Bertram Miller , C I and Thomas W. Beers (1982) 3rd edition “Forest Mensuration” John Wiley & Sons. Inc.
  • KFD (1977) working plan for sirsi and Siddapur Karnataka State Govt. publication.
  • FAO (1990) Remote sensing application to Land Resources. 14th UN/FAO Int. Trg, Course material.
  • SAC (1990″ Mannual of procedures for Forest Mapping and Damage Detection using satellite date ISRO – SAC-TR-88-90)
  • Howard, John A, (1991) “Remote Sensing of Forest Resources”. Chapman and Hall Edn.
  • P.S. Roy (1994) “Space Remote Sensing For Forest ECO_SYSTEM Analysis” Paper presented at 3rd ISY-Regional Remote Sensing Seminar on Tropical Eco-system management held during 23-25th August, 1994 at Bali (Indonesia)
  • Mohan Sundara Rajan (1991) “Remote Sensing and GIS for Natural Resources Management” ADB Manila.