Assessment and modelling of vegetation biomass in a major bauxite mine of Eastern Ghats, India

Abstract

Large-scale surface mining are the major causes for ecological disaster at the landscape level, but ecological restoration in post mining areas offers an opportunity to redevelop an ecosystem. The present research programme was undertaken in Panchapatmali Bauxite Mines in Koraput district of Odisha which is one of the biggest mines in Eastern Ghats ecoregion. The above ground biomass (AGB) and soil parameters inside and outside the mines (natural forest) were compared and their interrelationships were also tested at 1% level of significance. For the three dominant species Pinus insularis, Euca-lyptus hybrid and Samenia saman, the regression coefficient (R 2) values for AGB were significant with respect to DBH (R 2 = 0.80-0.90), height (R 2 = 0.35-0.76) and with basal area (R 2 = 0.90-0.96). Species wise, maximum biomass was shown by Pinus insularis followed by Eucalyptus hybrid and Samenia saman which reveals that exotic species have overruled the indigenous species in the plantation areas. An accurate cokriging geospatial model with minimum errors predicted the AGB values to range from 45.6 to 416.4 t/ha compared to the observed biomass range 5.90-507.06 t/ha through the developed regression equation y = 1.003x + 0.24. The overall AGB of the reclaimed area was at par with the natural forest outside the mines. Increasing the pH level of soils, planting indigenous species and increasing green ground cover species will have lesser negative competition with the trees in the reclaimed zone that can restore the fragile ecosystem.

Full text

1 23
Modeling Earth Systems and
Environment
ISSN 2363-6203
Model. Earth Syst. Environ.
DOI 10.1007/s40808-020-01004-4
Assessment and modelling of vegetation
biomass in a major bauxite mine of Eastern
Ghats, India
Kakoli Banerjee, Chandan Kumar Sahoo
& Rakesh Paul

1 23
Your article is protected by copyright and
all rights are held exclusively by Springer
Nature Switzerland AG. This e-offprint is
for personal use only and shall not be self-
archived in electronic repositories. If you wish
to self-archive your article, please use the
accepted manuscript version for posting on
your own website. You may further deposit
the accepted manuscript version in any
repository, provided it is only made publicly
available 12 months after official publication
or later and provided acknowledgement is
given to the original source of publication
and a link is inserted to the published article
on Springer's website. The link must be
accompanied by the following text: "The final
publication is available at link.springer.com”.

Vol.:(0123456789)
1 3
Modeling Earth Systems and Environment
https://doi.org/10.1007/s40808-020-01004-4
ORIGINAL ARTICLE
Assessment andmodelling ofvegetation biomass inamajor bauxite
mine ofEastern Ghats, India
KakoliBanerjee1 · ChandanKumarSahoo1· RakeshPaul1
Received: 13 August 2020 / Accepted: 8 October 2020
© Springer Nature Switzerland AG 2020
Abstract
Large-scale surface mining are the major causes for ecological disaster at the landscape level, but ecological restoration in
post mining areas offers an opportunity to re-develop an ecosystem. The present research programme was undertaken in
Panchapatmali Bauxite Mines in Koraput district of Odisha which is one of the biggest mines in Eastern Ghats ecoregion.
The above ground biomass (AGB) and soil parameters inside and outside the mines (natural forest) were compared and
their interrelationships were also tested at 1% level of significance. For the three dominant species Pinus insularis, Euca-
lyptus hybrid and Samenia saman, the regression coefficient (R2) values for AGB were significant with respect to DBH
(R2 = 0.80–0.90), height (R2 = 0.35–0.76) and with basal area (R2 = 0.90–0.96). Species wise, maximum biomass was shown
by Pinus insularis followed by Eucalyptus hybrid and Samenia saman which reveals that exotic species have overruled the
indigenous species in the plantation areas. An accurate cokriging geospatial model with minimum errors predicted the AGB
values to range from 45.6 to 416.4t/ha compared to the observed biomass range 5.90–507.06 t/ha through the developed
regression equation y = 1.003x + 0.24. The overall AGB of the reclaimed area was at par with the natural forest outside the
mines. Increasing the pH level of soils, planting indigenous species and increasing green ground cover species will have
lesser negative competition with the trees in the reclaimed zone that can restore the fragile ecosystem.
Keywords Eastern ghats· Bauxite mines· Reclamation· Above ground biomass· Soil parameters· Geospatial modelling
Introduction
Forest biomass and fields are renewable resources, which
provide alternative energy supply in rural areas (Bungart and
Huttal 2001). Heat production from different biomasses at
global scale has been projected to reach 864 million tonnes
of oil equivalent (TOE) by 2040 from an estimated 364
million TOE recorded in 2013 (IEA 2017; Nakahara etal.
2019). With the adaptation of National Renewable Energy
Act (NREA), 2015, the growth rate of bioenergy resources
has accelerated to promote the production of energy, so that
the dependence on fossil fuels can be minimized compensat-
ing with the socio-economic and environmental conditions.
According to estimations, 25% of the net energy utilisation
is met from biomass resources such as fuelwoods and other
forestry wastes (Kumar etal. 2002; Purohit and Fischer
2014). The renewable energy sources act as the means of
economic development in a developing country like India
and substitution of fossil-based materials with forest-based
materials to reduce the sources of greenhouse gases that can
potentially reduce the impact of climate change (Kilpelainen
etal. 2016).
Vast areas on our earth particularly the land resources are
exploited to the maximum by human activities (Choi and
Weali 1995). Such conditions are alarming because of the
fact that these changes are accelerating exponentially (Par-
rotta etal. 1997). Quarrying of minerals to meet the demand
of industries is compulsory (Bradshaw 1983). Human greed
for utilization of natural resources is the major causes for
destruction of biological diversity leading to various envi-
ronmental problems. Mining involves ecosystem degrada-
tions which are replaced by plantations in the overburden
dump (Piha etal. 1995). During the extraction of minerals
during mining process, the physical and biological nature
of mining area is drastically altered (in terms of soil quality,
soil texture, and biological productivity).
* Kakoli Banerjee
banerjee.kakoli@yahoo.com
1 Department ofBiodiversity andConservation ofNatural
Resources, Central University ofOdisha, Landiguda,
Koraput764021, Odisha, India
Author's personal copy

Modeling Earth Systems and Environment
1 3
Plantation of tree species is a challenging task as it is
affected by the changing soil and water parameters in the
plantation sites (Richardson and Greenwood 1967), undu-
lating topography (Brierley 1956; Down 1975) and com-
paction (Hall 1957; Richardson 1975). Sparse vegetation
in mining soils is basically due to reduced levels of organic
matter and plant nutrients such as phosphorous (Davison
and Jefferies 1966; Fitter and Bradshaw 1974), nitrogen
(Schramm 1966; Williams 1975), and potassium concen-
tration (Chadwick 1973) and high levels of metals alu-
minium (Al) and manganese (Mn) (Sutton and Dick 1987;
Kost etal. 1997). Most of the vegetation in mining areas
should be prevented from acidic soils (Sutton and Dick
1987). During mining, the OB dump areas are exposed
to iron disulphide (FeS2 i.e. pyrite) (Hill 1978). In the
process, the pyrites are exposed to atmospheric moisture
content, producing various acids and soluble salts (Singer
1970). The hydrogen (H+) ions generated reacts with Al,
Mn, Fe, Zn and Cu producing various metal oxides, car-
bonates, sulphate etc. (Massey and Barnhisel 1972; Barn-
hisel and Rotromel 1974). This is the reason why poorly
reclaimed or un-reclaimed lands are often barren as they
do not support plant growth.
Wide-spread destruction and degradation of natural habi-
tats has called for conservation aspects and CSR (corporate
social responsibility) for large-scale ecosystem restoration.
It is the process of restoring the diversity and dynamics of
indigenous ecosystem from damage (Jackson etal. 1995).
Afforestation programmes involving the regeneration of the
primary as well as secondary succession on plants lead to
nutrient generation in the OB dump areas helping in the
eco-restoration process (Dowarah etal. 2009). The lat-
eritic soils which covers about 10% of the total geographi-
cal area of India, are particularly poor in fertility with the
lack of nitrogen (N), phosphorus (P), and potassium (K)
and mining operations which keeps the soil exposed to air
exaggerate the problem (Bhat and Sujatha 2014; Dwevedi
etal. 2017). Hence, more amounts of nitrogen (N) and soil
organic carbon (OC) are of utmost importance, because
these nutrients limit the growth of trees as has also been
recorded in Jarrah forest (Abbott and Loneragan 1986) and
on reclaimed bauxite mines (Ward etal. 1985). Degraded
land can be reclaimed through regeneration of biotic factors
and increasing productivity (Brown and Lugo 1994). Hence,
long-term reclamation may lead to establishment of stable
nutrient cycle from plant growth. Afforestation methods are
in practice since time immemorial due to damage of forest
from anthropogenic factors (Filcheva etal. 2000). In addi-
tion, it also prevents soil erosion, increases SOM (Gill etal.
1987; Montagnini and Sancho 1990), decreases bulk den-
sity, moderates soil pH and nutrients (Sanchez etal. 1985;
Chakraborty and Chakraborty 1989; Sharma and Gupta
1989).
Tree biomass can be estimated through destructive,
non-destructive method and through allometric equations
(Rajashekar etal. 2018). Along with biomass, site-specific
influences of nutrients, temperature and precipitation on the
carbon assimilation are inevitable for assessment of primary
productivity which acts as the source-sink of CO2 (Knapp
and Smith 2001; Raich etal. 2006; Pan etal. 2011). Remote
sensing integrated with the ground data has become one of
the most effective methods for estimation of tree biomass
as it can be implemented in various topographic as well
as micro-climatic conditions of mountainous terrains with
better synoptic coverage and time-series analysis (Ravin-
dranath and Ostwald 2008; Singh etal. 2012; Salunkhe etal.
2016; Gwal etal. 2020). Keeping in view the topographic
and climatic conditions of a micro-region and using some
of the important predictors like vegetation indices (VIs)
extracted from the satellite imageries, geospatial models
have been used worldwide and in India in both tropical and
temperate conditions for the prediction of above ground
biomass (AGB) (Manna etal. 2014; Upgupta etal. 2015;
Gwal etal. 2020). Within a geographically and statistically
complex framework like Geographical Information System
(GIS) softwares, spatial predictive models like cokriging
have merged since the past 6decades which considers the
covariance factors for the interpolation of AGB in a closed
geographical area (Matheron 1963; Babcock etal. 2018).
Cokriging modelling in the GIS environment coupled with
ground inventory data and satellite-derived information,
improves the spatial prediction capabilities of vegetation
biomass compared to the techniques such as simple regres-
sion and ordinary kriging which does not take autocorrela-
tion into consideration resulting in over or under estimation
of the vegetation biomass (Mutanga and Rugege 2006; Singh
and Das 2014).
Normalised Difference Vegetation Index (NDVI) is pres-
ently in wide use in geospatial techniques for measuring pro-
ductivity of ecosystems (Pettorelli etal. 2005). The NDVI is
preferred in biomass modelling approaches as it reduces the
unwanted topographic, illuminative and atmospheric noises
along with the cloud cover errors present in the raw Land-
sat satellite imageries (Huete 2012). Previous works have
already been done on describing the quality of vegetation
using NDVI (Sinclair etal. 1971; Tucker 1979; Lobo etal.
1998; Aragon and Oesterheld 2008), to explore the function-
ing of ecosystem (Lloyd 1990; Reed etal. 1994; Hunt etal.
1996; Mysterud etal. 2007) and alterations recorded due to
global climatic changes (Penuelas and Filella 2001; Gong
and Ho 2003; Guo etal. 2008). Studies on inter-relationship
between NDVI and biomass parameters have resulted sig-
nificant differences with respect to different areas and NDVI
measures (Goward etal. 1985; Tucker etal. 1985; Box etal.
1989; Hobbs 1995; Gilabert etal. 1996; Schino etal. 2003).
This has led to the use of NDVI as the predictor variable in
Author's personal copy

Modeling Earth Systems and Environment
1 3
cokriging modelling and also linked with the ground truth
data in the current study. Based on the working hypothesis
that the reclaimed/plantation area has been eco-restored and
the vegetation has the AGB at par with the natural vegetation
outside the mines, the present study has aimed at the follow-
ing research questions: (1) what is the above ground biomass
(AGB) of the reclamation sites in the green belt zone of the
mines? (2) What is the status of soil parameters (soil pH, N,
P, K, OC and soil texture) of the mining area and their effects
on growth of the trees? (3) What is the biomass potential
of the reclaimed vegetation (inside mines) compared to the
natural vegetation (outer natural forest) through geospatial
modelling.
Materials andmethods
Site description
Koraput district has 98.82% of the state’s total production
of bauxite. Panchpatmali Bauxite Mines of Koraput Dis-
trict, Odisha falls partly in the Survey of India toposheet
65J/13 and partly in 65N/1. Here bauxite occurs on the
elongated plateau top as a blanket deposit which is about
21km long trending NNE–SSW with width varying from
1.8km to as low as 100m and elevation varies from 1200
to 1350m above MSL (Singh 2014). The plateau lies at an
average height of 300m above surrounding valley area. The
temperature during winter is as low as 3°C and in summer
rise to 40°C. Average rainfall is 1600–1800mm and wind
speed is 15km/h (usually in SW–NE direction), respectively.
Five (05) sites were selected namely Site1: below 10years;
Site2: 10–20years; Site3: 20–30years; Site4: natural forest
patch (in buffer zone) and Site5: barren land (Fig.1). The
first three Sites were taken in reclaimed area inside the mine
and the last two Sites were taken outside the mining area.
Vegetation structure, composition andeld
biomass analysis
Five quadrates/plots were taken in each sampling loca-
tion (25 plots in 5 sampling locations in total). Speci-
fied random sampling method was adopted in a 0.1ha
(31.62m × 31.62m) plot for estimation of vegetation
structure, composition and biomass of the trees. The sites
were selected based on the plantation map of the mines
provided by NALCO (National Aluminium Company),
Damanjodi of Koraput. The tree species were identified
with the help of the book Flora of Orissa by Saxena and
Brahmam (1997). Relative abundance was calculated to
understand the composition of the vegetation in the study
area which in turn helped to find out the biomass of three
(03) most dominant species and was calculated as per the
formula given by Achacoso etal. (2016):
where ni is the number of individuals of a particular species
and N is the total number of individuals including all species.
Non-destructive method was followed for vegetation bio-
mass analysis. Tree height, diameter at breast height (DBH)
(1.37m above the ground), basal area, and specific gravity
of wood were measured to calculate the biomass of each
species. The height of the tree was measured using Bosch
Range Finder DLE 40 (professional) instrument. The stem
volume of each tree was measured using the volume equa-
tion of a cylinder (considering the stem as a cylinder):
where V is the volume of the plant, π is the 3.14, r is the radius
of the plant and H is the height of the plant and f is the cor-
rection factor.
Wood specific gravities (g) for individual species were
determined by weighing an oven-dried stem wood core
of 1cm3 (Chaturvedi etal. 2010) which were finally used
for the estimation of AGB as per the expression given by
Brown and Lugo (1992):
(1)
Relative abundance
(RA)=
n
i
N
×
100,
(2)
V=
𝜋
r2fH,
Fig. 1 Map of the study Site
Author's personal copy

Modeling Earth Systems and Environment
1 3
where B is the biomass, g is the specific gravity and V is
the volume of stem.
Tree basal area (BA) was also calculated to establish its
correlation with the stem biomass as it has been proved to
be the most effective in predicting biomass over various leaf
area indices and tree canopy cover (Torres and Lovett 2013).
It is the cross-sectional area at the breast height which is
expressed in m2 and measured by the following expression
(Sahu etal. 2016):
where π is 3.14 and r is the radius.
Analysis ofsoil parameters
The soil samples were collected from five sites for the soil
physico-chemical parameter analysis in 1m × 1m quadrate
size. Before taking sample, all the litter and gravel were
cleared from the quadrate and soil was collected in an air
tight polythene bag from surface and sub-surface of depth
1–5cm from ground. The soil pH was measured by the help
of digital soil pH meter (SYSTRONICS) during field study.
The soil EC was measured by the help of digital EC meter
(Model no. EC testr). A mean of five readings was taken as
final. The soil samples were oven dried and sieved through
a 2mm sieve to retain the fine soil fractions (< 2mm) for
nutrient analysis.
Soil nitrogen was analysed through automatic distillation
system (Model Classic DX). For estimation, 5g of prepared
soil sample was taken in a digestion tube which was loaded
in distillation unit. Then, 25ml both of KMnO4 (0.32%) and
NaOH (2.5%) solution were added by distillation unit pro-
gramme. The sample mixture was heated through a flowing
steam and the ammonia liberated was absorbed in 20ml of
2% boric acid containing mixed indicator solution. It was
then kept in a 250ml conical flask. With the absorption of
ammonia, the solution turned from pink to green. About
150ml of distillate mixture was titrated with 0.02N H2SO4
which changed its colour to pink again. Simultaneously,
blank sample (without soil) was run. Then, both blank and
sample was titrated and the available nitrogen in soil was
calculated using following formula.
where R is the titration reading–blank reading, N is t he nor-
mality of acid and A is the atomic weight of nitrogen.
(3)
B=V×g,
(4)
BA =
𝜋
×r2,
(5)
Available nitrogen (
kg ha−1
)
=
R×N×A×
weight in 1ha
Sample weight (g)
×
1000 ,
Soil phosphorous was determined by Olsen’s method
(1954). 2.5 gm of air-dried sample was taken in 125ml
Erlenmeyer flask and a little of phosphorus free Darco-G
was added to it. Then, 50ml of NaHCO3 solution was added
at 25°C and was shaken for 30min on a reciprocating shaker
at 120 strokes/min. A blank (without soil) was also prepared
in the same procedure. From here, 10ml of aliquot was taken
in a 50ml volumetric flask through pipette and 10ml of de-
ionized water and one drop of p-nitrophenol was added to
it. Then, the whole content was acidified to pH 5.0 till the
colour disappears. Then, 8ml of the Murphy-Riley solution
was added and the volume was made up to 50ml with de-
ionized water. After 15min, the intensity of blue colour was
measured by spectrophotometer at 730nm.
where C is μg phosphorous in the aliquot (obtained standard
curve).
The standard flame photometry analysis was used for the
estimation of soil potassium. For this, 5 gm air-dried soil
sample was taken in a 150ml Erlenmeyer flask and then
25ml (1:5 soils to extractant) of neutral normal ammonium
acetate was poured. It was then shaken on a mechanical
shaker for about 5min and immediately filtered through
Whatman filter paper. Then, the reading was taken flame
and amount of K was calculated in the following formula.
where R is the flame photometer reading.
Soil Organic Carbon (OC) was analysed by method out-
lined by Walkley and Black (1934). For soil organic carbon
estimation, 1g of dried soil was taken in 500ml conical
flask to it 1ml of (H3PO4) and 1ml distilled water was
added. The mixture was heated for 10min at 100 to 110°C.
10ml 1N K2Cr2O7 and 20ml concentrated H2SO4 with
Ag2SO4 was added, mixed and allowed to stand for 30min.
The mixture was then diluted to 200ml and 10ml H3PO4
and 1ml freshly prepared standard solution was added to the
mixture till the colour of the mixture changed to bluish pur-
ple. Then, the mixture was titrated with Mohr salt solution
until the colour of the solution changed to brilliant green.
For blank, the same titration was repeated without taking
soil and the volume of K2Cr2O7 require to oxidised organic
carbon was calculated from the difference as follows:
where S is the volume of Mohr salt solution consumed by
sample, B is the volume of Mohr salt solution consumed by
blank and g is the weight of soil in grams.
(6)
Available phosphorus (
kg ha−1
)
=C×
Volume of extraction
Volume of aliquot
×
2.24
Weight of soil taken,
(7)
Available potassium
(K)
(
kg ha
−1)
=R×
11.217,
(8)
Soil organic carbon
(%)=
3.951
g
(1−
S
B
)
,
Author's personal copy

Modeling Earth Systems and Environment
1 3
For soil texture about 500g of soil sample were ana-
lysed for gravel, sand and silt/clay with the standard pro-
cedure of Sieve Shaker taste. The sieves were used to seg-
regate the amount of gravel, sand and silt/clay following
the grain scale as given by United States Department of
Agriculture (USDA) for classification of soil particle size.
Geospatial modelling forprediction
ofaboveground biomass
In total, 272 individual tree occurrences were taken in the
25 sampling plots used for the biomass prediction both
inside and buffer zone of the mining lease area to address
the research question that whether the reclaimed zone can
yield wood biomass at par with the natural vegetation
area outside the mines. The age factor of the plants was
not taken into consideration in the modelling to get more
continuous prediction in the study area. Landsat-8 OLI
(Operational Land Imager) imagery of 30m resolution
procured from the Earth Explorer archives (https ://earth
explo rer.usgs.gov) (tile-1, year-2018, month-December)
was used for the purpose of vegetation analysis as well as
biomass modelling.
We have chosen the cokriging model for the prediction
of biomass in the study area as the model performance
in cokriging encompassing the variogram models with
the ground biomass data increases considerably with bet-
ter accuracy as compared to the ordinary kriging and has
been outlined by various researchers as the better geospa-
tial models for AGB prediction (Papritz and Stein 1999;
Mutanga and Rugege 2006). This modelling technique
used multiple data columns in the attribute to find out the
cross-correlation and finally map with the semi-variogram
fit which lowers the mean-square prediction error (Singh
and Das 2014). Toposheet (scale: 1:50,000) procured from
SOI (Survey of India) was used for digitization of the
study area boundary and also for the accuracy assessment.
An expanded polygon was digitized covering enough area
outside the mining area to capture the natural vegetation
lying outside and draw a fair comparison of their predicted
and modelled biomass with the reclaimed zone inside the
mine. ArcMap 10.2.1 was used for the mapping and mod-
elling purpose.
The Normalised Differential Vegetation Index (NDVI)
was subsequently taken as the single predictor variable for
the cokriging model as it has been described as the best
predictor of above ground biomass among all the vegeta-
tion indices (Das and Singh 2016; Gwal etal. 2020). The
NDVI was calculated from the satellite imagery using the
Image Analysis tool available in ArcMap 10.2.1. NDVI was
computed using the expression given by Piao etal. (2007)
as follows:
where RNIR and RRED are the reflectance values of near-
infrared and red bands of the electromagnetic spectrum,
respectively.
To run the cokriging model, the vegetated area from the
non-vegetated area were separated in case of both inside and
outside the mining area. For this, Maximum Likelihood Clas-
sification (Lillesand etal. 2014) technique was adopted to
delimit the vegetation and non-vegetation areas. The aerial
imageries from Google Earth and the field sampling locations
were used for the accuracy assessment purpose using Kappa
statistics. As NDVI was taken as the predictor of AGB in the
current cokriging model, the correlation between the quadrat
wise field AGB data and the corresponding NDVI values using
scatter plot was established both inside and outside the mines
to ensure the strong positive correlation to increases the accu-
racy of the prediction model (Singh and Das 2014). Ordinary
cokriging was performed using the Box–Cox transformation
under exponential kernel function to plot the semi-variogram
of normally distributed dataset and also because of the stronger
spatial autocorrelation of the field AGB data. The spatial auto-
correlation of the field vegetation biomass was determined
using the geostatistical tool of ArcMap through the evalua-
tion of z-score and Moran’s Index (I) with the critical values
of 2.58 and − 0.003, respectively (Chen 2013). If the computed
values were more than the permissible level, the AGB can be
considered as spatially autocorrelated. The model nugget and
still were also determined in the output semi-variogram graph
where the former should be much closer to zero and the lat-
ter is the threshold at which the model curve becomes steady
(Yadav and Nandi 2015).
The accuracy of the cokriging model for the prediction
of above ground biomass was evaluated using the prediction
errors such as the mean error, root mean square error and root
mean square standardized error with minimum lag which finds
out the deviation of the modelled/predicted AGB from the field
AGB. The errors are calculated as per the given expressions
(Li etal. 2014; Singh and Das 2014):
(9)
NDVI
=
R
NIR
−R
RED
RNIR
+
RRED
,
(10)
Mean error
(ME)=1
N
N
∑
i=1
(̂e−e)
,
(11)
Root mean square error
(RMSE)=




1
N
N

i=1
(̂e−e)2
,
(12)
Root mean square standardized error
(RMSSE)=




1
N
N

i=1
̂e−e
e2
,
Author's personal copy

Modeling Earth Systems and Environment
1 3
where N is the sample size, ê is the biomass predicted, e is
the biomass observed and ē is the mean value of biomass.
When the mean error is near to zero (0) and the root
mean square standardized error is nearer to 1, it shows
the best fit model (Yadav and Nandi 2015). The complete
process of the geospatial analysis has been described in
the flow diagram (Fig.2). A final regression equation
was developed between the measured and predicted bio-
mass in the study area. For the comparison of the over-
all biomass potential inside and outside the mines, two
separate maps were generated for both the cases and the
modelled biomasses were extracted to point shapefiles in
ArcMap using the extract values to points feature. Finally,
they were compared in order to know the overall biomass
potential of the reclaimed plantation zone compared to
the natural vegetated area outside the mines. The final
modelled maps were generated with the contours of the
NDVI values in the ArcGIS interface in the study area for
better understanding and interpretation of the AGB values
with the NDVI values.
Statistical analysis
Pearson’s correlation coefficient (r) was calculated to find
the interrelationship between soil parameters and vegeta-
tion biomass to understand the effect of soil parameters on
growth of the planted species. Regression equation plots
were computed for selected species to prove the relationship
of AGB on height, DBH and basal area of the trees. One-way
ANOVA was computed to find the spatial difference in all
the selected parameters all data are entered in mean ± SD
and analysis were done using SPSS 22 (Figs.3, 4, 5).
Results anddiscussion
Restoration of mining sites is mainly done through adop-
tion of plantation in the overburdened sites to restore it
to natural ecosystem. This often helps in alteration of
physico-chemical characteristics of ambient environment,
thus restoring vegetation cover (Bradshaw 1987; Schaller
Fig. 2 Flow diagram of the Geospatial analysis used in the study
Author's personal copy

Modeling Earth Systems and Environment
1 3
1993). One of the important steps for ecological restora-
tion is by planting pioneer species enhancing the process
of succession so that ecosystem balance can be maintained
through biogeochemical cycles (Parrotta 1992). In this
matrix, the vegetation survey of five Sites in Panchpat-
mali bauxite mines were undertaken to analyse growth
parameters of trees with respect to biomass, soil quality
(in terms of N, P, K and OC) and the relationship between
soil parameter, vegetation biomass and NDVI which are
explained below.
Relative abundance ofplant species
This study revealed that Acacia auriculiformis, Eucalyp-
tus hybrid and Grevillea robusta were the most dominant
species considering the plantation Sites (Sites 1, 2 and
3) and the natural forest patch (Site-4). Studies on rela-
tive abundance of plant species is of utmost importance
to understand the adaptability of the species to the envi-
ronment, particularly when the plantations are being done
in reclaimed soil (overburden site). Such restoration of
mining habitats can be done through plantation of species
which are hardier and have positive effects to improve soil
nutrient capacity (Singh 1996; Parrota etal. 1997).
Soil physico–chemical properties
Overburden sites are generally low fertility soils, where
the substrate develops on the basis of the plantations
done for reclamation and restoration; therefore, random
soil samples were analyzed to determine the physical and
chemical variations in the plantation and non-plantation
sites in order to signify the difference in soil characteris-
tics. In the present study, soil was analysed for pH, OC,
EC, N, P, K, gravel, sand and silt/clay percent in all the
five selected sites. The pH value ranged from 4.34 (Site-
5) to 6.06 (Site-4), OC values ranges from 0.80 (Site-5)
to 1.62% (Site-4), EC values ranged from 0.09 (Site-1)
to 0.16mS cm−1 (Site-4), N, P, K values ranged from
276kg/h (Site-5), 2644kg/h (Site-4), 15kg/h (Site-5) to
554kg/h (Site-4), respectively; gravel percentage ranged
from 3.28 (Site-2) to 56.48% (Site-5); sand percent var-
ies from 31.76% (Site-5) to 78.2% (Site-2) and the silt/
clay percent varied from 6.9 (Site-4) to 21.76% (Site-3)
(Table1). Soil pH showed significant positive relationship
Fig. 3 Interrelationship between biomass and DBH, height and basal area (Acacia auriculiformis Bent.)
Author's personal copy

Modeling Earth Systems and Environment
1 3
with EC (p < 0.01) OC (p < 0.01), N, P, K (p < 0.01)
proving that with increasing nutrient in the soil pH has
increased (Table1). In case of Site-4 (natural forest patch),
there was a significant positive relationship (p < 0.01) of
soil pH with EC, however, significant negative relationship
was observed with respect to soil nutrient (N, P, K and
OC) at one percent (1%) level of significance (Table2).
Soil EC showed significant positive relationship
(p < 0.01) with respect to soil nutrient (N, P, K and OC)
in case of plantation Site (Sites 1, 2 and 3), whereas it
was just the reverse in case of natural forest patch (Site-4)
(Table1) which might be due to the location of the natu-
ral forest patch in the valley area where the nutrient gets
washed out whenever there is rainfall and hence the nutri-
ent is unable to accumulate in the study Site. Soil nutrient
N, P, K and OC have shown significant positive relation-
ship with each other (p < 0.01) both at the plantation site
as well as in the natural forest patch which proves that with
increasing organic matter (microbial degradation of leaf lit-
ter and also soil fauna) the soil N, P, K load has increased
(Tables1 and 2). Soil nutrient in the natural forest patch
has shown significant negative relationship (p < 0.01) with
respect to N, P, K at the plantation Site proving the accept-
ance of sandy soil of the planted species. The soils of the
reclaimed sites are free from gravel in comparison to wild
natural forest patch and barren land, whereas sand percent-
age is comparatively higher in the plantation site than the
other two sites (Table1). In case of natural forest patch
soil nutrient N, P, K and OC has shown significant posi-
tive relationship (p < 0.01) with respect to gravel and sand
and significant negative relationship silt/clay (p < 0.05;
p < 0.01) (Tables1 and 2). This proves that the natural
forest patch being a natural forest is acclimatized to the
environment. Such variations of soil nutrient and soil
texture with respect to plantation and growth of trees has
also been studied widely by many researchers (Fettweis
etal. 2005; Nicolini and Topp 2005; Banning etal. 2008;
Banning etal. 2012). ANOVA results between sites show
significant variations with respect to all physico-chemical
parameters of soil (p < 0.01) except EC, which proves that
there is significant variation between the sites and hence
may be responsible for variation in growth of planted spe-
cies (Table3). The existence of acidic soil in the study
site may be because of contamination of the overburden
soil after mining with pyrite (FeS2) that later oxidised sul-
phuric acid when exposed to oxygen and water which has
resulted in high level of soil acidification. The increased
soil nutrient N, P, K and OC with respect to increased age
Fig. 4 Interrelationship between biomass and DBH, height and basal area (Eucalyptus hibrid Maiden)
Author's personal copy

Modeling Earth Systems and Environment
1 3
of the plants in the plantation site have also been stated by
Mohr etal. (2005). The increased litter content in Sites 2
and 3 and the level of the land comparison to the natural
forest patch which is located in the valleys have resulted
in high ’N’ content in the study area. Several studies have
shown that soil pH and exchangeable cat ions in the soil
can positively influence tree species composition (Norden
1994; Finzi etal. 1998; Marschner and Noble 2000). We
assumed that in our result major differences between sites
with respect to soil nutrients may be a possible outcome of
the reasons mentioned above.
Biomass estimation
Biomass constitutes a major source of energy for about
50% of the world population (Karekez and Kithyoma
2006). Biomass in trees accounts for renewable source
of energy representing proportion of rural energy sup-
ply (Hashiramoto 2007) in addition to it, forest biomass
also contributes to food, fodder and fuel and hence its
degradation leads to ecosystem exploitation (Rawat and
Natutiyal 1988). Forest biomass accounts for amount
of carbon stored from the atmosphere in the process of
Fig. 5 Interrelationship between biomass and DBH (a), height (b) and basal area (c) (Grevillea robusta A.Cunn. ex R.Br.)
Table 1 Site wise soil physico-chemical parameters
Sites pH EC (mScm−1) OC % N (kg/ha) P (kg/ha) K (kg/ha) Gravel % Sand % Silt/clay %
Site1 (below
10year)
5.43 ± 0.12 0.09 ± 0.02 0.89 ± 0.10 414 ± 60.50 20 ± 1.20 208 ± 48.50 6.48 ± 3.03 72.12 ± 5.07 21.4 ± 5.03
Site2 (10–20year) 5.61 ± 0.26 0.1 ± 0.01 0.92 ± 0.08 495 ± 90.10 36 ± 0.89 380 ± 30.61 3.28 ± 2.25 78.2 ± 2.13 18.52 ± 2.85
Site3 (20–30year) 5.79 ± 0.32 0.11 ± 0.02 1.24 ± 0.15 566 ± 68.30 45 ± 1.25 435 ± 50.20 4.78 ± 3.11 73.46 ± 6.50 21.76 ± 6.41
Site4 (natural patch) 6.6 ± 1.23 0.16 ± 0.02 1.62 ± 0.16 644 ± 50.32 57 ± 4.25 554 ± 93.20 29.78 ± 6.03 63.32 ± 4.22 6.9 ± 2.40
Site5 (Barren land) 4.44 ± 0.49 0.1 ± 0.01 0.8 ± 0.20 276 ± 20.69 25 ± 1.02 180 ± 10.15 56.48 ± 4.77 31.76 ± 6.13 11.76 ± 2.11
Author's personal copy

Modeling Earth Systems and Environment
1 3
photosynthesis (Brown etal. 1999). Organic matter in the
soil also accounts for carbon storage in the soil (Vashum
and Jayakumar 2012). Biomass is a function of tree den-
sity, height, DBH and basal area at any given location, as it
differs with sites, habitats, forest successional stage, com-
position of forest, species variability, etc. (Brunig 1983;
Joshi and Ghose 2014).
In the present study, number of trees, DBH, tree height,
basal area and biomass have been estimated in four selected
Sites with different age groups (Site-1: < 10years, Site-2:
10–20years, Site-3: 20–30years and Site-4: natural forest
patch, whose age is not defined as it is the natural forest).
The biomass in Site-1 varied as per the order G. robusta
(12.72 ± 0.03 t/ha) > S. cumini (7.20 ± 0.09 t/ha) > A.
auriculiformis (7.13 ± 0.17t/ha) > P. pinnata (1.04 ± 0.03
t/ha) > E. hibrid (0.69 ± 0.03t/ha), respectively (Table4),
at Site-2, the biomass of the tree species varied as per the
order G. robusta (202.92 ± 0.75 t/ha) > A. auriculiformis
(88.92 ± 0.58t/ha) > A. excels (7.36 ± 0.17t/ha) > C. equi-
setifolia (0.57 ± 0.05 t/ha) > P. guajava (0.05 ± 0.001t/
ha), respectively (Table 5), Site-3 the biomass varied as
per the order E. hybrid (555.39 ± 6.03 t/ha) > P. insularis
(471.96 ± 2.85 t/ha) > A. auriculiformis (85.41 ± 0.43 t/
ha) > S. saman (48.48 ± 3.54t/ha) > G. robusta (8.4 ± 0.53t/
ha) > C. fistula (5.1 ± 1.11 t/ha) > M. elengi (1.5 ± 0.16t/
ha) > B. purpurea (1.4 ± 0.70t/ha) > T. arjuna (1.32 ± 0.02t/
ha) > C. robusta (0.14 ± 0.003t/ha), respectively (Table6).
At Site-4, biomass values varied as per the order M. indica
(9.61 ± 0.48t/ha) > A. excels (6.67 ± 0.06t/ha) > M. indica
(6.240.002 t/ha) > F. glomerata (5.12 ± 0.001 t/ha) > P.
pinnata (4.25 ± 1.09t/ha) > S. anacardium (4.13 ± 0.02 t/
ha) > T. indica (3.92 ± 0.02t/ha) > T. chebula (3.17 ± 0.01t/
ha) > T. arjuna (3.05 ± 0.01t/ha) > C. collinus (2.35 ± 0.01t/
ha) > A. leucophloea (2.27 ± 0.03t/ha) > S. cumini (2.24t/
ha) > E. hybrid (2.07 ± 0.01t/ha) > Z. oenoplia (1.98 ± 0.04t/
ha) (Table7).
The overall study on biomass of 34 species highest bio-
mass was shown by P. insularis followed by E. hibrid, S.
saman. Site wise variation in biomass showed Site-3 to have
highest biomass in case of plantation plot which is basi-
cally due to the age of the plant. However, Site-4 the (natu-
ral forest patch) has shown more or less a higher biomass
values owing to its adaptability in the natural environment.
The overall site wise variation of biomass was of the order,
Site-3 (1179.10 ± 211.33 t/ha) > Site-4 (821.76 ± 66.33t/
Table 2 Interrelationship between AGB and soil parameters (Plantation site)
* p < 0.05; **p < 0.01
pH EC OC % N (kg/ha) P (kg/ha) K (kg/ha) Gravel % Sand % Silt/clay % Biomass t/ha
pH 1
EC 1** 1
OC % 0.90** 0.90** 1
N kg/ha 1.00** 1.00** 0.89** 1
P kg/ha 0.99** 0.99** 0.82** 0.99** 1
K kg/ha 0.96** 0.96** 0.74** 0.97** 0.99** 1
Gravel − 0.53* − 0.53* − 0.11 − 0.56** − 0.66** − 0.75** 1
Sand 0.21 0.21 − 0.23 0.25 0.36 0.48* − 0.94** 1
Silt/clay 0.10 0.10 0.52* 0.06 − 0.06 − 0.19 0.79** − 0.95** 1
Biomass t/ha 0.96** 0.96** 0.99** 0.94** 0.90** 0.83** − 0.26 − 0.09 0.39 1
Table 3 One way ANOVA
showing variation between the
selected sites
Parameters SS df MS F p value F crit
pH 12.07272 4 3.02 389.74 1.16E−18 2.87
EC mScm−1 0.01836 4 0.00 49.89 3.97E−10 2.87
OC (%) 2.233576 4 0.56 141.29 2.32E−14 2.87
N (kg/ha) 414,141.8 4 103,535.46 1090.31 4.3E−23 2.87
P (kg/ha) 4450.208 4 1112.55 99.02 6.93E−13 2.87
K (kg/ha) 496,294 4 124,073.50 83.58 3.43E−12 2.87
Gravel (%) 10,601.5 4 2650.39 159.50 7.00E−15 2.87
Sand (%) 6983 4 1745.86 68.27 2.00E−11 2.87
Silt/clay (%) 847.262 4 211.82 12.52 3.00E−05 2.87
Biomass (kg/ha) 202,570.9 4 50,642.7 5.4834 0.0038 2.8661
Author's personal copy

Modeling Earth Systems and Environment
1 3
Table 4 Total biomass (t/ha) of the tree species found at Site-1 (average of five quadrates)
Sl No. Scientific
name
Common
Name
No of trees CBH (m) DBH (m) Ht-1st-
forking
(m)
Tree-height
(m)
Radius (m) Correc-
tion factor
(f)
Basal Area Volume
(m3)
Wood
density
(kg/m3)
Biomass (t/
ha)
Total biomass
(t/ha)
1Acacia
auricu-
liformis
Bent
Acasia 31 0.31 0.1 1.88 5.14 0.05 0.26 0.009 0.045 520 0.23 7.13 ± 0.17
2Syzygium
cumini
(L.) Skeels
Jamu 60 0.25 0.08 0.77 2.94 0.04 0.2 0.006 0.018 650 0.12 7.2 ± 0.09
3Pongamia
pinnata
(L.) Pierre
Karanja 26 0.12 0.04 1.42 4.17 0.02 0.23 0.001 0.004 920 0.04 1.04 ± 0.03
4Eucalyptus
hibrid
Maiden
Nilagiri 23 0.14 0.04 0.7 2.06 0.02 0.22 0.002 0.004 910 0.03 0.69 ± 0.03
5Grevillea
robusta
A.Cunn.
ex R.Br
Silver oak 318 0.17 0.05 1.05 2.76 0.03 0.26 0.003 0.008 510 0.04 12.72 ± 0.04
Author's personal copy

Modeling Earth Systems and Environment
1 3
ha) > Site-2 (299.82 ± 88.26t/ha) > Site-1 (28.50 ± 5.01 t/
ha).
Correlation coefficient computed for biomass in the plan-
tation site shows significant positive relationship (p < 0.01)
with pH, EC, OC and N, P, K which proves that the planta-
tion Site is very well nurtured for the planted species. The
less amount of gravel and sand in the area has also sup-
ported the growth of the species (Table8) similar significant
positive relationship (p < 0.01) of biomass with OC, EC, N,
P, K, gravel and sand proves that the tree species found in
the natural environment are well accustomed to the natural
environment for years together. However, significant nega-
tive relationship (p < 0.01) with pH, EC and silt/clay speaks
that with decrease in pH and salt concentration the biomass
has increased (Table8). ANOVA constituted for biomass
between Sites also showed significant variation between
Sites (p < 0.05) which clearly speaks that with increase in
age the biomass of the trees increases (Table3).
Regression equations were plotted for AGB with height,
DBH and basal area for selected tree (A. auriculiformis, G.
robusta and E. hibrid) on the basis of relative abundance in
the study sites. This was basically done to understand the
relationship of height, DBH, and basal area to predict the
increase in biomass through non-destructive approach. For
all the species, the R2 values for AGB were significant with
respect to DBH (R2 = 0.80–0.90), height (R2 = 0.35–0.76)
and with basal area (R2 = 0.90–0.96). Similar studies on rela-
tionship of DBH, height and basal area of above ground bio-
mass (AGB) has also been done widely (Tumwebaze etal.
2013; Owate etal. 2018).
Biomass prediction modelling
The NDVI values ranged from a minimum of − 0.351 to a
maximum of 0.404 inside the mining area and from − 0.056
to 0.50 outside the mining area. The NDVI values of the
vegetation outside the mines recorded comparatively higher
than the vegetation in the reclaimed sites inside mines show-
ing the slightly better growth and vigour of the vegetation in
the non-mining area compared to the reclaimed zone where
mining activity was previously carried out (Xue and Su
2017). The reason for the higher values may be attributed
to the higher age of the plants in the natural area than the
plantation sites and the undisturbed environment for their
growth. However, the difference in the NDVI values inside
and outside the mines is very less and comparatively at par
with each other which tells us that the reclaimed areas have
not lost or regained their potential and resilience to support
tree growth and ultimately produce more wood biomass. The
maximum likelihood classification was performed to distin-
guish the vegetated and non-vegetated areas inside and out-
side the mines with an accuracy of 97.65% with the Kappa
Coefficient of 9.53 proving that the vegetated area taken for
Table 5 Total biomass (t/ha) of the tree species found at Site-2 (average of five quadrates)
Sl No Scientific
Name
Common
Name
No of trees CBH (m) DBH (m) Ht-1st-
forking
(m)
Tree-Height
(m)
Radius (m) Correc-
tion factor
(f)
Basal Area Volume
(m2)
Wood
density
(kg/m3)
Biomass (t/
ha)
Total biomass
(t/ha)
1Acacia
auricu-
liformis
Bent
Acacia 342 0.27 0.09 2.6 5.75 0.04 0.31 0.007 0.0504 520 0.26 88.92 ± 0.58
2Casuarina
equisetifo-
lia L
Jhanu 3 0.22 0.07 1.33 6.07 0.03 0.15 0.004 0.0229 830 0.19 0.57 ± 0.05
3Ailanthus
excelsa
Roxb
Mahanimba 23 0.29 0.09 2.05 5.03 0.05 0.27 0.007 0.0371 860 0.32 7.36 ± 0.17
4Psidium
guajava L
Pijuli 1 0.18 0.06 1.3 3.2 0.03 0.27 0.003 0.0083 590 0.05 0.05 ± 0.001
5Grevillea
robusta
A.Cunn.
ex R.Br
Silver oak 356 0.34 0.11 3.08 6.93 0.05 0.31 0.012 0.1114 510 0.57 202.92 ± 0.75
Author's personal copy

Modeling Earth Systems and Environment
1 3
Table 6 Total Biomass (t/ha) of the tree species found at Site-3 (average of five quadrates)
Sl No. Scientific
name
Common
name
No of trees CBH (m) DBH (m) Ht-1st-
forking
(m)
Tree-height
(m)
Radius (m) Correc-
tion factor
(f)
Basal Area Volume
(m2)
Wood
density
(kg/m3)
Biomass (t/
ha)
Total biomass
(t/ha)
1Acacia
auricu-
liformis
Bent
Acacia 219 0.32 0.1 2.61 6.25 0.05 0.29 0.01 0.0742 520 0.39 85.41 ± 0.43
2Terminalia
arjuna
(Roxb.
ex DC.)
Wight &
Arn
Arjuna 4 0.38 0.12 2.1 4.43 0.06 0.32 0.012 0.0544 610 0.33 1.32 ± 0.02
3Samanea
saman L
Bada
Chakunda
24 0.43 0.14 4.77 11.58 0.07 0.31 0.024 0.4697 430 2.02 48.48 ± 3.45
4Mimusops
elengi L
Baula 5 0.29 0.09 1.54 5.48 0.05 0.2 0.007 0.0413 720 0.3 1.5 ± 0.16
5Coffea
robusta
L.Linden
Coffee 7 0.11 0.03 1.3 2.04 0.02 0.43 0.001 0.0018 1090 0.02 0.14 ± 0.003
6Bauhinia
purpurea
L
Kanchana 2 0.34 0.11 5 8.75 0.05 0.37 0.01 0.0969 720 0.7 1.4 ± 0.70
7Eucalyptus
hibrid
Maiden
Nilagiri 99 0.56 0.18 5.58 13.64 0.09 0.28 0.033 0.6166 910 5.61 555.39 ± 6.03
8Pinus insu-
laris L
Pine 138 0.63 0.2 11.06 18.51 0.1 0.4 0.037 0.7277 470 3.42 471.96 ± 2.85
9Grevillea
robusta
A.Cunn.
ex R.Br
Silver oak 14 0.27 0.09 6.52 14.75 0.04 0.32 0.007 0.1169 510 0.6 8.4 ± 0.53
10 Cassia
fistula L
Sunari 6 0.34 0.11 3.77 8.82 0.05 0.29 0.01 0.1195 710 0.85 5.1 ± 1.11
Author's personal copy

Modeling Earth Systems and Environment
1 3
Table 7 Total biomass (t/ha) of the tree species found at Site-4 (average of five quadrates)
Sl No. Scientific
name
Common
name
No of trees CBH (m) DBH (m) Ht-1st-
forking
(m)
Tree-height
(m)
Radius (m) Correc-
tion factor
(f)
Basal Area Volume
(m2)
Wood
density
(kg/m3)
Biomass (t/
ha)
Total biomass
(t/ha)
1Mangifera
indica L
Amba 28 2.8 0.89 4.2 18.3 0.45 0.15 0.020 1.75 550 9.61 269.16 ± 0.48
2Eucalyptus
hibrid
Maiden
Nilagiri 54 0.9 0.29 5.3 18.1 0.14 0.20 0.001 0.23 910 2.07 111.97 ± 0.01
3Ficus glom-
erata. L
Dimiri 8 1.95 0.62 6.5 21.5 0.31 0.20 0.005 1.31 390 5.12 40.93 ± 0.001
4Pongamia
pinnata
(L.) Pierre
Karanja 7 1.1 0.35 7.2 19.2 0.18 0.25 0.003 0.46 920 4.25 29.78 ± 1.09
5Tamarindus
indica L
Tentuli 9 1.55 0.49 4.1 19.3 0.25 0.14 0.010 0.52 750 3.92 35.29 ± 0.02
6Ziziphus
oenoplia L
Kanta koli 9 0.85 0.27 6.8 10.8 0.14 0.42 0.005 0.26 760 1.98 17.84 ± 0.04
7Acacia leu-
cophloea
(Roxb.)
Willd
Gohira 8 0.93 0.30 6.5 15.5 0.15 0.28 0.013 0.30 760 2.27 18.14 ± 0.03
8Syzygium
cumini (L.)
Skeels
Jamu 9 1.05 0.33 5.9 12.2 0.17 0.32 0.005 0.35 650 2.24 20.20 ± 0.01
9Terminalia
arjuna
(Roxb.
ex DC.)
Wight &
Arn
Arjuna 13 1.12 0.36 7.5 18.8 0.18 0.27 0.004 0.50 610 3.05 39.60 ± 0.01
10 Madhuca
indica
J.F.Gmel
Mahula 11 1.87 0.60 4.1 20.8 0.30 0.13 0.008 0.76 820 6.24 68.64 ± 0.02
11 Terminalia
chebula
Retz
Harida 8 1.15 0.37 6.1 20.8 0.18 0.20 0.006 0.43 740 3.17 25.35 ± 0.01
12 Cleistanthus
collinus
Benth.
exHook.f
Karada 10 0.92 0.29 6.3 17.8 0.15 0.24 0.008 0.28 830 2.35 23.49 ± 0.01
13 Semecarpus
anacar-
diumL.f
Bhalia 10 1.08 0.34 7.1 13.2 0.17 0.36 0.007 0.44 940 4.13 41.32 ± 0.02
Author's personal copy

Modeling Earth Systems and Environment
1 3
the cokriging biomass prediction model was accurate. The
field biomass and corresponding NDVI values showed a sig-
nificant positive correlation both within and outside the min-
ing area with the values of R2 = 0.99 and R2 = 0.94, respec-
tively, which ensured the correct interpolation of the wood
biomass in the study area (Singh and Das 2014) (Fig.6a, b).
The model nugget was found at zero which shows the
presence of a complete and strong spatial structure and
the still was observed at 4425.8 at the range of 889.4m
(Fig.7) in the cross-variogram analysis suggesting that the
cross-variogram attained stability after a number of itera-
tions proving the current model to be a fit model (Yadav
and Nandi 2015). The spatial auto-correlation performed
for the biomass sampled data which showed a z value of
62.43 (Critical value = 2.58) and Moran’s Index (I) of 0.82
(expected index = − 0.003). This showed that the biomass
data used for the modelling were spatially autocorrelated
and the average high and low values were more clus-
tered. The final regression equation between the meas-
ured and predicted values of biomass was established as
“y = 1.003x + 0.24” where “y” is the above ground biomass
(t/ha) and “x” is the NDVI value, which can be further
used on the satellite imageries of the nearby areas of the
mine in Koraput district to get the predicted biomass map
(Fig.8a). The number of lags was found to be 12 in the
model which was calculated using the nearest neighbour-
hood toolbox in ArcGIS and is the acceptable threshold
for an accurate cokriging modelling (Singh and Das 2014).
In the cross-validation process of the cokriging model,
the mean error, root mean square error and root mean
square standardized errors were found to be − 0.111,
9.156Mgha−1 and 0.63, respectively (Fig.8b). Here, the
values of mean error and root mean square standardized
error are nearer to the optimal values (0 and 1, respec-
tively) showing an excellent fit of the model. The meas-
ured values of biomass ranged from 5.90 to 507.06t/ha,
whereas the predicted biomass values ranged from 45.6
to 416.4t/ha in the total study area (inside and outside
the mine) (Figs.8a, 9 and 10). The reclaimed area has
surprisingly shown the predicted above ground biomass
from the range 45.6 to 402.1 t/ha which was at par with
the natural vegetation areas outside the mine showing the
values from 47.3 to 416.4t/ha (Figs.9 and 10). The reason
of the reverting biomass potential of the mining area can
be linked to the fact that it is very less prone to grazing
activities and also adjacent to the natural forest present
outside the mines (Franklin etal. 2012; Macdonald etal.
2015). The current findings can be utilised by the mining
authorities in consultation with the forest department of
the district for eco-restoration of the mines in the districts.
Some of the studies carried out in deciduous forests also
suggest the technique of rooting the rough surfaces which
would support the recolonization and growth of trees and
Table 7 (continued)
Sl No. Scientific
name
Common
name
No of trees CBH (m) DBH (m) Ht-1st-
forking
(m)
Tree-height
(m)
Radius (m) Correc-
tion factor
(f)
Basal Area Volume
(m2)
Wood
density
(kg/m3)
Biomass (t/
ha)
Total biomass
(t/ha)
14 Ailanthus
excelsa
Roxb
Mahanimba 12 1.63 0.52 5.5 18.5 0.26 0.20 0.010 0.78 860 6.67 80.05 ± 0.06
Author's personal copy

Modeling Earth Systems and Environment
1 3
Table 8 Interrelationship between AGB and soil parameters (natural forest patch)
* p < 0.05; **p < 0.01
pH EC OC % N (kg/ha) P (kg/ha) K (kg/ha) Gravel % Sand % Silt/clay % Biomass t/ha
pH 1
EC 0.85** 1
OC % − 0.92** − 0.99** 1
N kg/ha − 0.96** − 0.96** 0.99** 1
P kg/ha − 1.00** − 0.89** 0.95** 0.98** 1
K kg/ha − 0.88** − 0.50* 0.64** 0.72** 0.84** 1
Gravel − 1.00** − 0.82** 0.90** 0.94** 0.99** 0.91** 1
Sand − 0.98** − 0.74** 0.84** 0.90** 0.96** 0.95** 0.99** 1
Silt/clay 0.72 0.24 − 0.40 − 0.50 − 0.65 − 0.96 − 0.76 − 0.83 1
Biomass t/ha − 0.99 − 0.90 0.96 0.98 1.00 0.83 0.99 0.96 − 0.64 1
Fig. 6 Interrelationship between AGB and NDVI inside (a) and outside (b) the mining area
Fig. 7 Semivariogram showing Nugget and still values of model
Author's personal copy

Modeling Earth Systems and Environment
1 3
also considering the plantation of ground cover species
which will have lesser negative competition with the trees
in the reclaimed zone (Zipper etal. 2011; Macdonald etal.
2015).
Conclusion
Enumeration of forest biomass is a relevant parameter for
understanding ecosystem balance and climate change stud-
ies, as well as for forest management. In the present study,
the satellite spectral information (NDVI) has shown a good
correlation with AGB combined with field measurement,
which has been proven through cokriging modelling. This
Geostatistical technique has been useful in interpolating the
AGB to unsampled locations to give a continuous and holis-
tic picture of the biomass potential inside the outside the
mining area. Semi-variogram and cross-variogram plotted
and observed in cokriging model proved it to be the best fit
exponential model providing a stable cross-variogram and
variance between the measured and predicted variables as
the mean error and RMSE were within the optimal level. The
predicted AGB values of the adjacent natural forest being
almost equal to that of the reclaimed mining area tell us
the effort of the district forest department and the mining
authority towards the eco-restoration process.
The study suggests that addition of lime in the planta-
tion Site as well as in barren land will increase the soil
pH which will help in the growth of the plants. Addition
of organic fertiliser in the form of vegetable compost or
vermi-compost will help in increasing OC and NPK in the
soil. The barren land can be planted with pioneer species
like lemon grass (Cymbopogan citratus) will increase the
fertility of the bare soil. Plantation of indegenous spe-
cies like Sal (S. robusta), Sisoo (D. sisoo), Teak (T. gran-
dis), Bhalia (S. anacardium), Arjuna (T. arjuna), Karanja
(P. pinnata), Harida (T. chebula), Bahada (T. bellerica),
Chandan (S. album) will help in restoring the biodiver-
sity of Koraput of Odisha. Continuous monitoring of bio-
mass through field vegetation data along with geospatial
techniques [NDVI, k-NN, CoK, DRR (Direct Radiometric
radiation)] will help to understand the health of ecosystem
from time to time.
For the stakeholders, this study will help in enhance-
ment of floral biodiversity in the reclaimed area. Increase
in green cover will help in reduction of respiratory dis-
eases that occur due to increased suspended particulate
matter (SPM) and elevated carbon dioxide levels in mining
Fig. 8 Regression equation of measured and predicted values of biomass with error
Author's personal copy

Modeling Earth Systems and Environment
1 3
areas. As per MoEF and CC, Government of India guide-
lines, the mines have to plant over equivalent non-forest
land that is degraded during mining. For this, selected
indigenous species should be planted during Compensa-
tory Afforestation (CA) programme to maintain the green
belt of the mines.
Acknowledgements The authors duly acknowledge the financial sup-
port received from NRSC Hyderabad project (Sanction No. NRSC/
FEG/VCP dt 03/07/2015), constant support of GM (Mines) NALCO,
Damanjodi for providing appropriate permissions in the field. The sec-
ond author is grateful to UGC for providing Non-Net fellowship and
last author is grateful to DST-INSPIRE for providing fellowship (Sanc-
tion No. DST/INSPIRE Fellowship/2015/IF150127 dated 10.04.2015)
for undertaking the research.
Funding The authors duly acknowledge the financial support received
from NRSC Hyderabad project (Sanction No. NRSC/FEG/VCP dt
03/07/2015).
Availability of data and material The data should be made available
on authors’ permission.
Compliance with ethical standards
Conflict of interest The authors declare that they have no conflict of
interest.
Code availability No custom codes were utilized in the research.
References
Abbott I, Loneragan O (1986) Ecology of Jarrah (Eucalyptus margi-
nata) in the northern Jarrah forest of Western Australia Bulle-
tin No. 1. Department of Conservation and Land Management
Perth, WA, p 137
Achacoso SC, Walag AMP, Saab LL (2016) A rapid assessment of
foliage spider fauna diversity in Sinaloc, El Salvador City,
Philippines: a comparison between habitats receiving differ-
ent degrees of disturbance. Biodiversity 17:156–161. https ://
doi.org/10.1080/14888 386.2016.12583 31
Aragon R, Oesterheld M (2008) Linking vegetation heterogene-
ity and functional attributes of temperate grasslands through
remote sensing. Appl Veg Sci 11(1):117–130. https ://doi.
org/10.1111/j.1654-109X.2008.tb002 10.x
Babcock C, Finley AO, Andersen HE, Pattison R, Cook BD, Morton
DC, Alonzo M, Nelson R, Gregoire T, Ene L, Gobakken T,
Næsset E (2018) Geostatistical estimation of forest biomass in
interior Alaska combining Landsat-derived tree cover, sampled
Fig. 9 Predicted map of Biomass-NDVI inside the mine
Fig. 10 Predicted map of Biomass-NDVI outside side the mine
Author's personal copy

Modeling Earth Systems and Environment
1 3
airborne lidar and field observations. Remote Sens Environ
212:212–230. https ://doi.org/10.1016/j.rse.2018.04.044
Banning NC, Grant CD, Jones DL, Murphy DV (2008) Recovery
of soil organic matter; organic matter turn over and nitrogen
cycling in a post- mining forest rehabilitation chronosequence.
Soil Biol Biochem 40(8):2021–2031. https ://doi.org/10.1016/j.
soilb io.2008.04.010
Banninga NC, Lalora BM, Cooksona WR, Griggb AH, Murphya
DV (2012) Analysis of soil microbial community level physi-
ological profiles in native and post-mining rehabilitation forest:
which substrates discriminate? Appl Soil Ecol 56:27–34. https
://doi.org/10.1016/j.apsoi l.2012.01.009
Barnhisel RI, Rotromel AL (1974) Weathering of clay minerals by
simulated acid coal spoil-bank solutions. Soil Sci 118(1):22–27
Bhat R, Sujatha S (2014) Soil fertility status and disorders in are-
canut (Areca Catechu L.) grown on clay and laterite soils of
India. Commun Soil Sci Plant Anal 45:1622–1635. https ://doi.
org/10.1080/00103 624.2014.90791 0
Box E, Holben B, Kalb V (1989) Accuracy of the AVHRR vegetation
index as a predictor of biomass, primary productivity and net
CO2 flux. Vegetatio 80:71–89
Bradshaw AD (1983) The reconstruction of ecosystems. J Appl Ecol
20:1–17
Bradshaw AD (1987) Restoration ecology. In: Jordan WR, Gilpin ME,
Aber JD (eds) A synchronised approach to ecological restoration.
Cambridge University Press, Cambridge, pp 61–74
Brierley JK (1956) Some preliminary observations on the ecology of
pit heaps. J Ecol 44:383–390
Brown S, Lugo AE (1992) Aboveground biomass estimates for tropi-
cal moist forest of the Brazilian Amazon. Interciencia 17:8–18
Brown S, Lugo AE (1994) Rehabilitation of tropical lands: a key to
sustaining development. Restor Ecol 2(2):97–111. https ://doi.
org/10.1111/j.1526-100X.1994.tb000 47.x
Brown SL, Schroeder P, Kern JS (1999) Spatial distribution of biomass
in forests of the eastern USA. Forest Ecol Manag 123(1):81–90.
https ://doi.org/10.1016/S0378 -1127(99)00017 -1
Brunig EF (1983) Structure and growth. In: Golley FB (ed) Ecosystems
of the World 14 A, Tropical rain forest ecosystems: structure and
function. Elsevier Scientific Publication, New York, pp 49–75
Bungart R, Huttl RF (2001) Production of biomass for energy in post-
mining landscapes and nutrient dynamics. Biomass Bioenerg
20(3):181–187. https ://doi.org/10.1016/S0961 -9534(00)00078 -7
Chadwick MJ (1973) Methods of assessment of acid colliery spoil as a
medium for plant growth. In: Hutnik RJ, Davis G (eds) Ecology
and reclamation of devastated land, vol 1. Gordon and Breach
Publishing Group, London, pp 81–91
Chakraborty RN, Chakraborty D (1989) Changes in soil properties
under Acacia auriculiformis plantation in Tripura. Indian For
115(4):272–273
Chaturvedi RK, Raghubanshi AS, Singh JS (2010) Non-destructive
estimation of tree biomass by using wood specific gravity in the
estimator. Natl Acad Sci Lett 33:133–138
Chen Y (2013) New approaches for calculating Moran’s Index of spa-
tial autocorrelation. PLoS ONE. https ://doi.org/10.1371/journ
al.pone.00683 36
Choi YD, Weali MK (1995) The role of Panicum virgatum (Switch
Grass) in the revegetation of iron-mine tailings in Northern New
York. Restor Ecol 3(2):123–132. https ://doi.org/10.1111/j.1526-
100X.1995.tb000 85.x
Das S, Singh TP (2016) Forest type, diversity and biomass estimation
in tropical forests of western ghat of maharashtra using geospa-
tial techniques. Small-Scale For 18(62):517–532. https ://doi.
org/10.1007/s1184 2-016-9337-y
Davison A, Jefferies BJ (1966) Some experiment on the nutrition of
plants growing on coal mine waste heaps. Nature 210:649–650
Dowarah J, Deka Boruah HP, Gogoi J, Pathak N, Saikia N, Handique
AK (2009) Eco-restoration of a high-sulphur coal mine overbur-
den dumping site in northeast India: a case study. J Earth Syst
Sci 118:597–608. https ://doi.org/10.1007/s1204 0-009-0042-5
Down CG (1975) Soil development on colliery waste tips in rela-
tion to age. I. Introduction and physical factors. J Appl Ecol
12(2):613–622
Dwevedi A, Kumar P, Kumar P, Kumar Y, Sharma YK, Kayastha AM
(2017) Soil sensors: detailed insight into research updates, sig-
nificance, and future prospects. In: Grumezescu AM (ed) New
pesticides and soil sensors. Elsevier Inc, p 566. http://dx.doi.
org/https ://doi.org/10.1016/B978-0-12-80429 9-1.00016 -3
Fettweis U, Bens O, Huttl RF (2005) Accumulation and properties of
soil organic carbon at reclaimed sites in the Lusatian lignite min-
ing district afforested with Pinus sp. Geoderma 129(1–2):81–91.
https ://doi.org/10.1016/j.geode rma.2004.12.034
Filcheva E, Noustorova M, Sv KG, Haigh MJ (2000) Impact of foresta-
tion on organic accumulation and microbial action in surface coal
mine-spoil Pernik Bulgaria. Ecol Eng 15(1):1–15. https ://doi.
org/10.1016/S0925 -8574(99)00008 -7
Finzi AC, Canham CD, Van Breemen N (1998) Canopy tree- soil
interactions within temperate forests: species effects on pH and
cations. Ecol Appl 8(2):447–454. https ://doi.org/10.1890/1051-
0761(1998)008[0447:CTSIW T]2.0.CO;2
Fitter AH, Bradshaw AD (1974) Root penetration of Lolium perenne
on colliery shale in response to reclamation treatment. J Appl
Ecol 11(2):609
Franklin O, Johansson J, Dewar RC, Dieckmann U, McMurtrie RE,
Brännström AK, Dybzinski R (2012) Modeling carbon allocation
in trees: a search for principles. Tree Physiol 32:648–666. https
://doi.org/10.1093/treep hys/tpr13 8
Gilabert M, Gandia S, Melia J (1996) Analyses of spectral-biophysical
relationships for a corn canopy. Remote Sens Environ 55(1):11–
20. https ://doi.org/10.1016/0034-4257(95)00187 -5
Gill HS, Abrol IP, Samra JS (1987) Nutrient recycling through litter
production in young plantations of Acacia nilotica and Euca-
lyptus tereticornis in a highly alkaline soil. Forest Ecol Manag
22(1–2):57–69. https ://doi.org/10.1016/0378-1127(87)90095 -8
Gong D, Ho C (2003) Detection of large-scale climate signals in spring
vegetation index (Normalized Difference Vegetation Index) over
the Northern Hemisphere. J Geophys Res 108(D16):1–12. https
://doi.org/10.1029/2002J D0023 00
Goward S, Tucker C, Dye D (1985) North American vegetation pat-
terns observed with the NOAA-7 advanced very high resolution
radiometer. Vegetatio 64(1):3–14. https ://doi.org/10.1007/BF000
33449
Guo W, Yang T, Dai J, Shi L, Lu Z (2008) Vegetation covers changes
and their relationship to climate variation in the source region
of the Yellow River, China, 1990–2000. Int J Remote Sens
29(7):2085–2103. https ://doi.org/10.1080/01431 16070 13952 29
Gwal S, Singh S, Gupta S, Anand S (2020) Understanding forest bio-
mass and net primary productivity in Himalayan ecosystem using
geospatial approach. Model Earth Syst Environ 6:2517–2534.
https ://doi.org/10.1007/s4080 8-020-00844 -4
Hall IG (1957) The ecology of disused pit help in England. J Ecol
45(3):689–720
Hashiramoto O (2007) Wood-product trade and policy issue. In: Dubé
Y (ed) Cross-sectoral policy developments in forestry. CABI
Wallingford Publication, Wallingford, pp 24–35
Hill RD (1978) Methods for controlling pollutions. In: Schaller FW,
Sutton P (eds) Reclamation of drastically disturbed lands. Ameri-
can Society of Agronomy, Madison, pp 687–704
Hobbs T (1995) The use of NOAA-AVHRR NDVI data to assess herb-
age production in the arid rangelands of Central Australia. Int J
Remote Sens 16(7):1289–1302. https ://doi.org/10.1080/01431
16950 89544 77
Author's personal copy

Modeling Earth Systems and Environment
1 3
https ://earth explo rer.usgs.gov. “Accessed on 15th January, 2019.
Huete AR (2012) Vegetation indices, remote sensing and forest
monitoring. Geogr Compass 6:513–532. https ://doi.org/10.111
1/j.1749-8198.2012.00507 .x
Hunt E, Piper S, Nemani R, Keeling C, Otto R, Running S (1996)
Global net carbon exchange and intra-annual atmospheric CO2
concentrations predicted by an ecosystem process model and
three-dimensional atmospheric transport model. Global Bio-
geochem Cy 10(3):431–456. https ://doi.org/10.1029/96GB0 1691
International Energy Agency: Key word energy statistics 2017.
International Energy Agency. https ://www.iea.org/publi catio
ns/freep ublic ation s/publi catio n/KeyWo rld20 17.pdf (2017).
Accessed 15 Jun 2018
Jackson LL, Lopoukhine N, Hillyard D (1995) Ecological restora-
tion: a definition and comments. Restor Ecol 3(2):71–75. https
://doi.org/10.1111/j.1526-100X.1995.tb000 79.x
Joshi HG, Ghose M (2014) Community structure, species diversity,
and aboveground biomass of Sundarbans Mangrove swamps.
Trop Ecol 55(3):283–303
Karekez S, Kithyoma W (2006) Bio-energy and agriculture: prom-
ises and challenges Bio-energy and the poor. In: 2020 Vision
for Food Agriculture and the Environment. International Food
P Res Institute, Washington
Kilpeläinen A, Torssonen P, Strandman H, Kellomäki S, Asikainen
A, Peltola H (2016) Net climate impacts of forest biomass
production and utilization in managed boreal forests. GCB Bio-
energy 8:307–316. https ://doi.org/10.1111/gcbb.12243
Knapp AK, Smith MD (2001) Variation among biomes in tem-
poral dynamics of aboveground primary production. Sci-
ence 80(291):481–484. https ://doi.org/10.1126/scien
ce.291.5503.481
Kost DA, Boutelle DA, Larson MM, Smith WD, Vimmerstedt JP
(1997) Paper mill sludge amendment, tree protection and
establishment on an abandoned coal mine soil. J Environ Qual
26(5):1409–1416. https ://doi.org/10.2134/jeq19 97.00472
42500 26000 50030 x
Kumar A, Purohit P, Rana S, Kandpal TC (2002) An approach to the
estimation of value of agricultural residues used as biofuels.
Biomass Bioenergy 22(3):195–203. https ://doi.org/10.1016/
S0961 -9534(01)00070 -8
Li W, Niu Z, Gao S, Wang C (2014) Mapping the spatial pattern of
temperate forest above ground biomass by integrating airborne
lidar with Radarsat-2 imagery via geostatistical models. Lidar
Remote Sens Environ Monit XIV 9262:92620S. https ://doi.
org/10.1117/12.20686 43
Lillesand TM, Kiefer RW, Chipman JW (2014) Remote sensing and
image interpretation, 7th edn. Wiley, New York, USA (ISBN
978-1-118-34328-9)
Lloyd D (1990) A phonological classification of terrestrial vegetation
cover using shortwave vegetation index imagery. Int J Remote
Sens 11(12):2269–2279. https ://doi.org/10.1080/01431 16900
89551 74
Lobo A, Moloney K, Chic O, Chiariello N (1998) Analysis of
finescale spatial pattern of a grassland from remotely–sensed
imagery and field collected data. Landsc Ecol 13(2):111–131.
https ://doi.org/10.1023/A:10079 38526 886
Macdonald SE, Landhäusser SM, Skousen J, Franklin J, Frouz J, Hall
S, Jacobs DF, Quideau S (2015) Forest restoration following
surface mining disturbance: challenges and solutions. New For
46:703–732. https ://doi.org/10.1007/s1105 6-015-9506-4
Manna S, Nandy S, Chanda A, Akhand A, Hazra S, Dadhawal VK
(2014) Estimating aboveground biomass in Avicennia marina
plantation in Indian Sundarbans using high-resolution satellite
data. J Appl Remote Sens 8:083638. https ://doi.org/10.1117/1.
jrs.8.08363 8
Marschner B, Noble AD (2000) Chemical and biological processes
leading to the neutralization of acidity in soil incubated with
litter material. Soil Biol Biochem 32(6):805–813. https ://doi.
org/10.1016/S0038 -0717(99)00209 -6
Massey HF, Barnhisel RI (1972) Copper, Nickel and zinc released
from acid coal mine spoil materials of eastern Kentucky. Soil
Sci 113(3):207–212
Matheron G (1963) Principles of geostatistics. Econ Geol
58:1246–1266
Mohr D, Simon M, Toppa W (2005) Stand composition affects
soil quality in oak stands on reclaimed and natural sites.
Geoderma 129(1–2):45–53. https ://doi.org/10.1016/j.geode
rma.2004.12.029
Montagnini F, Sancho F (1990) Impacts of native trees in tropical soils:
a study in the Atlantic lowlands of Costa Rica. Ambio 19(8):386–
390. https ://www.jstor .org/stabl e/43137 46
Mutanga O, Rugege D (2006) Integrating remote sensing and spa-
tial statistics to model herbaceous biomass distribution in a
tropical savanna. Int J Remote Sens 27:3499–3514. https ://doi.
org/10.1080/01431 16060 06397 35
Mysterud A, Tryjanowski P, Panek M, Pettorelli N, Stenseth N (2007)
Inter-specific synchrony of two contrasting ungulates: wild boar
(Sus scrofa) and roe deer (Capreolus capreolus). Oecologia
151(2):232–239. https ://doi.org/10.1007/s0044 2-006-0584-z
Nakahara Y, Tabata T, Ohno T, Furukawa F, Inokuchi K, Katagiri K,
Hirayama Y (2019) Discussion on regional revitalization using
woody biomass resources as renewable energy. Int J Energy Envi-
ron Eng 10:243–256. https ://doi.org/10.1007/s4009 5-019-0300-5
Nicolini F, Topp W (2005) Soil properties in plantations of sessile oak
(Quercus petraea) and red oak (Quercus rubra) in reclaimed lig-
nite open-cast mines of the Rhineland. Geoderma 129(1–2):65–
72. https ://doi.org/10.1016/j.geode rma.2004.12.031
Norden U (1994) Influence of tree species on acidification and mineral
pools in deciduous forest soils of South Sweden. Water Air Soil
Poll 76(3–4):363–381
Owate OA, Mware MJ, Kinyanjui MJ (2018) Allometric equations
for estimating Silk Oak (Grevillea robusta) biomass in agricul-
tural landscapes of Maragu Subcounty, Kenya. Int J Forest Res
2018:1–14. https ://doi.org/10.1155/2018/64952 71
Pan Y, Richard A, Pekka E etal (2011) A large and persistent carbon
sink in the world’s forests. Science 80(333):988–993. https ://doi.
org/10.1126/scien ce.12016 09
Papritz A, Stein A (1999) Spatial prediction by linear kriging. In: Stein
A, van der Meer F, Gorte B (eds) Spatial statistics for remote
sensing. Kluwer, Dordrecht, pp 83–113
Parrotta JA (1992) The role of plantation forests in rehabilitating
degraded tropical ecosystems. Agr Ecosyst Environ 41(2):115–
133. https ://doi.org/10.1016/0167-8809(92)90105 -K
Parrotta JA, Turnbullb JW, Jonesc N (1997) Catalyzing native for-
est regeneration on degraded tropical lands. Forest Ecol Manag
99(1–2):1–7. https ://doi.org/10.1016/S0378 -1127(97)00190 -4
Penuelas J, Filella I (2001) Phenology—responses to a warming world.
Science 294:793–795. https ://doi.org/10.1126/scien ce.10668 60
Pettorelli N, Vik J, Mysterud A, Gaillard JM, Tucker C, Stenseth
N (2005) Using the satellite-derived NDVI to assess eco-
logical responses to environmental change. Trends Ecol Evol
20(9):503–510
Piao SL, Fang JY, Zhou LM, Tan K, Tao S (2007) Changes in bio-
mass carbon stocks in China’s grasslands between 1982 and
1999. Glob Biogeochem Cycles 21(2):1–10. https ://doi.
org/10.1029/2005G B0026 34
Piha MI, Vallack HW, Michael N, Reeler BM (1995) A low input
approach to vegetation establishment on mine and coal ash
wastes in semi-arid regions. II. Lagooned pulverized fuel ash in
Zimbabwe. J Appl Ecol 32(2):382–390. https ://doi.org/10.1016/j.
tree.2005.05.011
Author's personal copy

Modeling Earth Systems and Environment
1 3
Purohit P, Fischer G (2014) Second-generation biofuel potential in
India: sustainability and cost considerations. UNEP Risø Centre
on Energy, Climate and Sustainable Development, Copenhagen
Raich JW, Russell AE, Kitayama K, Patron WJ, Vitousek PM (2006)
Temperature influences carbon accumulation in moist tropical
forests. Ecology 87:76–87. https ://doi.org/10.1890/05-0023
Rajashekar G, Fararoda R, Reddy RS, Jha CS, Ganeshaiah KN, Singh
JS, Dadhwal VK (2018) Spatial distribution of forest biomass
carbon (Above and below ground) in Indian forests. Ecol Indic
85:742–752. https ://doi.org/10.1016/j.ecoli nd.2017.11.024
Ravindranath NH, Ostwald M (2008) Carbon inventory methods
handbook for greenhouse gas inventory, carbon mitigation
and roundwood production projects. In: Advances in global
change research. Springer, Berlin. https://doi.org/https ://doi.
org/10.1007/978-1-4020-6547-7
Rawat JK, Nautiyal JC (1988) Forest biomass: a source of food, feed
and fuel. Indian For 114(8):429–439
Reed B, Brown J, Vanderzee D, Loveland T, Merchant J, Ohlen D
(1994) Measuring phenological variability from satellite imagery.
J Veg Sci 5(5):703–714. https ://doi.org/10.2307/32358 84
Richardson JA (1975) Physical problem of growing plant on colliery
waste. In: Chadwick MJ, Goodman GT (eds) The ecology of
resource degradation and renewal. Blackwell Scientific Publica-
tions, Oxford, pp 275–285
Richardson JA, Greenwood EF (1967) Soil moisture tension in relation
to plant colonisation of pit heaps. Proc Univ Newctle Tyne Phil
Soc 1:129–136
Rouse JW, Haas RH, Schell JA, Deering, DW (1974) Monitoring veg-
etation systems in the Great Plains with ERTS. In: Freden SC,
Mercanti EP, Becker M (eds) Third earth resources technology
satellite–1 syposium. Volume I: Technical Presentations, NASA
SP-351, NASA, Washington DC, pp 309–317
Sahu SC, Suresh HS, Ravindranath NH (2016) Forest structure, compo-
sition, and above ground biomass of tree community in tropical
dry forests of Eastern Ghats India. Not Sci Biol 8(1):125–133.
https ://doi.org/10.15835 /nsb81 9746
Salunkhe O, Khare PK, Sahu TR, Singh S (2016) Estimation of tree
biomass reserves in tropical deciduous forests of Central India by
non-destructive approach. Trop Ecol 57(2):153–161
Sanchez PA, Palm CA, Davey CB, Szott LT, Russell CE (1985) Trees
as soil improver in the humid tropics? In: Cannel MGR, Jackson
JE (eds) Attributes of trees as crop plants. Institute of Terrestrial
Ecology, Abbots Ripton, Huntingdon, pp 327–350
Saxena HO, Brahmam M (1997) The Flora of Orissa. Orissa Forest
Development Corporation Limited, Bhubaneswar, pp 1–4
Schaller N (1993) The concept of agricultural sustainability. Agr
Ecosyst Environ 46(1–4):89–97. https ://doi.org/10.1016/0167-
8809(93)90016 -I
Schino G, Borfecchia F, De Cecco L, Dibari C, Iannetta M, Martini S,
Pedrotti F (2003) Satellite estimate of grass biomass in a moun-
tainous range in central Italy. Agroforest Syst 59(2):157–162
Schramm JR (1966) Plant colonization studies on black waste from
anthracite mining in Pennsylvania. Trans Am Philos Soc
56(1):1–194. https ://doi.org/10.1023/A:10263 08928 874
Sharma BD, Gupta IC (1989) Effect of tree cover on soil fertility in
western Rajasthan. Indian For 115(5):348–354
Sinclair T, Hoffer R, Schreiber M (1971) Reflectance and internal
structure of leaves from several crops during a growing season.
Agron J 633(6):864–868
Singer PC, Stumm W (1970) Acidic mine drainage- the rate determin-
ing step. Science 167(3921):1121–1123
Singh TP, Das S (2014) Predictive analysis for vegetation biomass
assessment in Western Ghat Region (WG) using geospatial
techniques. J Indian Soc Remote Sens 42(3):549–557. https ://
doi.org/10.1007/s1252 4-013-0335-7
Singh JS, Singh KP, Jha AK (1996) Final Technical Report submitted
to the Ministry of Coal, Govt. of India, p 116
Singh S, Dadhwal VK, Patil P (2012) Assessment of aboveground
phytomass in temperate forests of Kashmir valley, J & K, India.
Int J Ecol Environ Sci 38:47–58
Sutton P, Dick WA (1987) Reclamation of acidic mined lands in humid
areas. Adv Agron 41:377–405. https ://doi.org/10.1016/S0065
-2113(08)60809 -3
Torres AB, Lovett JC (2013) Using basal area to estimate aboveground
carbon stocks in forests: La Primavera Biosphere’s Reserve,
Mexico. Forestry 86:267–281. https ://doi.org/10.1093/fores try/
cps08 4
Tucker C (1979) Red and photographic infra red linear combination
for monitoring vegetation. Remote Sens Environ 8(2):127–150.
https ://doi.org/10.1016/0034-4257(79)90013 -0
Tucker C, Vanpraet C, Sharman M, Vanittersum G (1985) Satel-
lite remote-sensing of total herbaceous biomass production
in the Senegalese Sahel—1980–1984. Remote Sens Environ
17(3):233–249. https ://doi.org/10.1016/0034-4257(85)90097 -5
Tumwebaze SB, Bevilacqua E, Briggs R, Volk T (2013) Allometric
biomass equations for tree species used in agroforestry sys-
tems in Uganda. Agroforestry Syst 87(4):781–795. https ://doi.
org/10.1007/s1045 7-013-9596-y
Upgupta S, Singh S, Tiwari PS (2015) Estimation of aboveground
phytomass of plantations using digital photogrammetry and
high resolution remote sensing data. J Indian Soc Remote Sens
43:311–323. https ://doi.org/10.1007/s1252 4-014-0401-9
Vashum KT, Jayakumar S (2012) Methods to estimate above ground
biomass and carbon stock in natural forest—a review. J Ecosyst
Ecogr 2(4):1–7. https ://doi.org/10.4172/2157-7625.10001 16
Walkley AJ, Black IA (1934) Estimation of soil organic carbon by the
chromic acid titration method. Soil Sci 37(1):29–38. https ://doi.
org/10.1097/00010 694-19340 1000-00003
Ward SG, Pickersgill GE, Michaelsen DV, Bell DT (1985) Responses
to factorial combinations of nitrogen, phosphorus and potassium
fertilizers by saplings of Eucalyptus saltgna Sm. and the predic-
tion of the responses by DRIS indices. Aust For Res 15(1):27–32
Williams PJ (1975) Investigation in to the nitrogen cycle in colliery
spoils. In: Chadwick MJ, Goodman GT (eds) The ecology of
resource degradation and renewal. Blackwell Scientific Publica-
tions, Oxford, pp 259–274
Xue J, Su B (2017) Significant remote sensing vegetation indices: a
review of developments and applications. J Sensors. https ://doi.
org/10.1155/2017/13536 91
Yadav BKV, Nandy S (2015) Mapping aboveground woody biomass
using forest inventory, remote sensing and geostatistical tech-
niques. Environ Monit Assess. https ://doi.org/10.1007/s1066
1-015-4551-1
Zipper CE, Burger JA, Skousen JG, Angel PN, Barton CD, Davis V,
Franklin JA (2011) Restoring forests and associated ecosystem
services on Appalachian coal surface mines. Environ Manag
47:751–765. https ://doi.org/10.1007/s0026 7-011-9670-z
Publisher’s Note Springer Nature remains neutral with regard to
jurisdictional claims in published maps and institutional affiliations.
Author's personal copy