Use of Current Remote Sensing Methods for Biodiversity Monitoring

and Conservation of Mount Kilimanjaro National Park Ecosystems

Fortunata Msoffe

1,2

, Thomas Nauss

and Dirk Zeuss

Philipps Universitat, Umweltinformatik, Deutschhausstrasse, 12, 35032, Marburg-Hessen, Germany

Tanzania National Parks, P.O. Box 3134, Arusha, Tanzania

Keywords: Climate-land-use Change, Kilimanjaro-Mountain National Park, Remote Sensing, Sentinel-2,

Vegetation-indices.

Abstract: Climate and land use change have become serious challenges facing protected areas globally, more so those

in the tropical forest ecosystems. Kilimanjaro-Mountain National Park was specifically designated to protect

and safeguard the highest free-standing mountain in the tropics. The park attracts thousands of national and

international tourists annually because of its snow capped-summit and the altitudinal gradients, representing

the different eco-climatic zones of the world. Earnings from tourism boost the country’s economy while

ensuring the sustainability of this unique glacial-tropical mountainous forest ecosystem park. Conventional

monitoring of key biodiversity and environmental parameters are carried out by park staff, following

established guidelines by Tanzania National Parks. However, given the park’s geo-morphological nature of

mountainous terrain, efficient implementations of the labor intensive in-situ observations are hardly

feasible. This research explored the use of Remote Sensing data from the European Satellite Agency–

Sentinel- 2 Multi-Spectral Instrument, in developing a state-of-the-art monitoring protocol. The developed

methodology ensures that essential biodiversity parameters, including Vegetation Indices, required in

monitoring the vast areas of the park and its surroundings in the short-term and long-term, using up to date,

high resolutions and frequently available Remote Sensing data from the Sentinel-2 Sensors are captured.

1 INRODUCTION

It is well known that protected areas particularly

those in the tropics face key challenges linked to loss

of wildlife habitats, mainly due to land use changes

in their surrounding ecosystems whilst exacerbated

by the increasing impacts of global climate change

(Peters, et al. 2016; Burgess, et al., 2017; Tabor, et

al., 2018). Kilimanjaro Mountain National Park and

its associated ecosystem represents such a world-

wide unique and diverse habitat, with an altitudinal

range of over 5,800 m and associated with climate

and vegetation zones changing from the tropical

savannas at the lowlands to the afro-alpine

grasslands at the top (Hemp, 2009). Apart from the

natural ecosystems within the national park, several

land use types occur in the vicinity, including

intensive annual monocultures (maize and other

cereals), perennial coffee-plantations and diverse

traditional agro-forestry systems of Chagga-Home-

gardens which to some extent retain a semi-natural

forestry structure (Hemp and Hemp, 2018).

This highest free-standing mountain in Africa

acts as a water tower by feeding major river systems

in the region. The tropical mountainous forest

ecosystem plays a major role in the regional climate

regulation, while providing many other important

ecosystem services to the locals and beyond (Hemp

and Hemp, 2018). Its melting “ice cap”, which is an

important tourism attraction by mountaineers and

tourists visiting the park every year, though caused

by decreasing precipitation (Hemp, 2005;

Thompson, et al. 2002) rather than by increasing

temperature, has become a global symbol for the

accelerating trend of global warming. The KiLi-

Project, funded by the German Research Foundation

(DFG) studied the influence of climate and land-use

change on biodiversity and multiple ecosystems

processes on Mount Kilimanjaro from 65 established

plots, across twelve different land covers and land

uses along the elevation gradient (vertically from the

lowlands of Colline savannas to the highest peak of

vegetation layer dominated by the Hellichrysum)

and across the land use gradient from the protected

Msoffe, F., Nauss, T. and Zeuss, D.

Use of Current Remote Sensing Methods for Biodiversity Monitoring and Conservation of Mount Kilimanjaro National Park Ecosystems.

DOI: 10.5220/0009357701750183

In Proceedings of the 6th International Conference on Geographical Information Systems Theory, Applications and Management (GISTAM 2020), pages 175-183

ISBN: 978-989-758-425-1

175

tropical montane forest of Kilimanjaro National Park

to the disturbed lowlands of the savannas currently

converted to intensive monoculture crops cultivation

(Appelhans, et al., 2016).

This study, capitalizes on the recently concluded

“KiLi1”-Project (2010-2018), with the main

objective of a follow-up monitoring strategy for the

Kilimanjaro National Park, being the custodian in

ensuring the continuity of the ecosystem services

provided by the park to the local, national and the

international community at large. Apart from the

direct ecosystem services provided by the park in its

natural settings, it is particularly a key tourist

destination in the country, contributing to the local

and national economy from the foreign currency

accrued through the tourism business and its tripling

effects to the local communities surrounding the

park. In doing so, Kilimanjaro National Park is

vested with the responsibility of protecting this

uniquely massif standing tropical montane cloud

forest in the long run, at the face of its increasingly

isolation from its surroundings, mainly through

habitat conversions from the natural forest

vegetation to croplands because of the adjacent

intensifying land uses spearheaded by the increasing

human population pressure (Hemp and Hemp,

2018).

The study explored the use of remote sensing by

deploying the current state of the art from the

Sentinel-2 Multi-Spectral Instrument (MSI) satellite

of the European Satellite Agency (ESA), in

developing a workflow protocol data-model tool.

The implementation of the workflow protocol will

enable “in-situ repeated observations up-scale”,

which are hardly feasible in such a large protected

area’s challenging ecosystem by park staff. The

currently available data from the Sentinel-2 MSI

provides multi-spectral bands with high spatial

resolutions and quick revisit time of five days for

both sentinel 2 A & B (ESA, 2017). The Sentinel- 2

MSI is comprised of 13 spectral bands ranging in

resolutions from 10 m, (four bands) inclusive of the

visible wavelengths; 20 m (six bands) inclusive of

the new “Red-Edge”, near-infra red and short-wave

infra-red wavelengths; important for vegetation

monitoring and with high capabilities for use in

ecosystem, biodiversity and conservation monitoring

(Drusch, et al. 2012). The other three bands are of 60

m resolution including the aerosol, water vapor and

cirrus bands (Table 1).

Table 1: Spectral bands available from the Sentinel- 2 A

(since June 2015) and Sentinel- 2 B (since March 2017)

NIR = near infra-red; SWIR = short wave infra-red;

(Credit: ESA, 2017).

Sentinel -2 MSI

Bands

Central

Wavelengths

(µm)

Resolution

(in m)

Band 1 – Coastal

Aerosol

0.443 60

Band 2 – Blue 0.490 10

Band 3 - Green 0.560 10

Band 4 – Red 0.665 10

Band 5-Red edge 1 0.705 20

Band 6-Red edge 2 0.740 20

Band 7-Red edge 3 0.783 20

Band 8- NIR 0.842 10

Band 8A – NIR 0.865 20

Band 9-Water vapor 0.945 60

Band 10 – SWIR-

Cirrus

1.375 60

Band 11-SWIR- 1 1.610 20

Band 12-SWIR- 2 2.190 20

Spectral signatures and derived indices like the

Normalized Difference Vegetation Index (NDVI)

are used as a standardized way to measure the health

of vegetation by quantifying the ratio of the

difference between the NIR (strongly reflected by

vegetation) and Red (strongly absorbed by

vegetation). NDVI values ranges from -1 to + 1,

with a distinct boundary for each type of land cover.

Negatives likely represent water, while positives

close to one indicate dense green leaves. However,

values close to zeros represent no leaves (green

vegetation) or degraded forest (Figure 1). In this

research project, Sentinel – 2 MSI spectral bands

and derived products, including the vegetation and

biophysical indices such as NDVI and leaf area

index (LAI), were explored and analyzed with the

following objectives;

• Explore current RS data and methods in

developing a biodiversity and conservation

monitoring tool in the Kilimanjaro National

Park and its surrounding ecosystems

• Link remote sensing derived information

with in-situ data and predict/retrieve

spatially explicit biodiversity measures

• Demonstrate the explanatory power in

biodiversity analysis and its contribution to

a deeper understanding of the biodiversity

GISTAM 2020 - 6th International Conference on Geographical Information Systems Theory, Applications and Management

176

and potential linkages to ecosystem

functions along climatic and anthropogenic

disturbance gradients at Mt. Kilimanjaro.

Figure 1: Spectral Bands Reflectance’s Mapping

Application of the Sentinel-2 MSI (Credit: ESA, 2017).

2 MATERIALS AND METHODS

2.1 Study Area

The Kilimanjaro Mountain National Park and its

ecosystems, is located in the north-east of Tanzania

(Figure 2) and spans an elevation gradient from the

colline savanna plains (~ 700 m a.s.l.) to the

glaciated areas encircling Kibo summit (5895 m

a.s.l.). Its equatorial daytime climate is shaped by

the passing of the inter-tropical convergence zone,

with more than half of the annual rainfall occurring

during the so called long-rainy season (March-May),

(Appelhans, et al., 2016). While annual precipitation

amounts to more than 2500 mm in the southern

montane forest belt, the northern mountain side (lee

ward) receives hardly more than 1000 mm (Hemp

2005; Detsch, et al., 2017).

The mountain’s belt-like vegetation zonation

(Figure 3) is characterized by major land-cover

transitions at short horizontal distances resulting

from changing climatic conditions and

anthropogenic interferences (Hemp & Hemp 2018).

This study covered the land-cover distributional

zones which were marked by the 65 Kili-Research

Project Plots (Detsch, et al., 2017; Appelhans, et al.

2016), which were distributed along the mountain

elevation gradient; from the lowlands of savanna

vegetation to the top elevation zone covered by the

Hellichrysum spp. Plots were also selected to cover

twelve land-cover uses, from the total protected

forests in the park, dominated by Ocotea,

Podocarpus and Helichrysum vegetation to the

intensive monoculture crop cultivation dominated by

Maize and the Coffee plantations.



Figure 2: Study Area- the Kilimanjaro Mountain Ecosystems in Africa-Tanzania-(inset), showing the distribution of the

sampled plots along the elevation and land-cover/use types gradient (Appelhans, et al., 2016).

Use of Current Remote Sensing Methods for Biodiversity Monitoring and Conservation of Mount Kilimanjaro National Park Ecosystems

177

Figure 3: Map of Kilimanjaro Mountain Ecosystems showing land-cover/use vegetation types (Hemp and Hemp, 2018).

2.2 Sentinel 2 MSI Data Sets

Sentinel- 2 datasets used in this project were

downloaded from the ESA-Copernicus website,

https://sentinels.copernicus.eu/web/sentinel/sentinel-

data-access/registration. The

Kilimanjaro is covered

with two-sentinel 2 tiles area wide; the T37MBS on

the western side and the T37MCS on the eastern part

according to the date of acquisition, in this case

scenes taken between 2016 and 2019. Sentinel 2 A

and B image data were searched and downloaded for

the study area. The selected scenes were chosen

based on cloud cover percentages in order to obtain

good quality images. However, given the nature of

the Kilimanjaro, being a tropical cloud montane

forest, it is almost impossible to acquire cloud-free

scenes at any time of the year. Images with less than

10 percent of cloud covers were downloaded from

both sentinel 2 A and B, as Level 1 C Top of

Atmosphere (TOA) products, through the Sentinel

Application Platform (SNAP) Software. SNAP is

also available for free through downloading at the

ESA-Copernicus website, with its associated plug-

ins, for the pre-processing of the Sentinel-2 data.

The downloaded images were further processed

for atmospheric effects from the top of atmosphere

Level 1C to bottom of atmosphere (BOA) Level 2A

products, using the Sen2Cor Algorithm-plug-in from

SNAP (Wilm, 2018). In order to automate the

process, the image products (L1C) were exported

from the SNAP- graphical processing tool (.gpt) into

the R software, as .xml files for commands and

functions creation using the R Studio processing

environment. In R, image files were further

corrected for atmospheric effects, to Level 2A

products using the Sen2Cor algorithm. Image

outputs were then used for the follow-up stages of

extracting the various vegetation and biophysical

indices from the sentinel-2 imagery data. In order to

create a cloudless image, the T37-MBS and T37-

MCS tiles had to be processed separately before

mosaicking of the tiles using the Cloud-Mask layer,

also produced as part of the level 2A processing

output, performed in R as shown in Figure4below.

GISTAM 2020 - 6th International Conference on Geographical Information Systems Theory, Applications and Management

178

Figure 4: Workflows for access, pre- and processing,

analyses and sharing Outputs of Image Data from

Sentinel-2 through the TANAPA GIS Server.

2.3 Vegetation and Biophysical Indices

Derived from Sentinel-2 Data

Creation of commands-data models for indices

extraction automation was performed after the

atmospheric corrections for each of the scenes,

according to the date of acquisition, as year, month

and day for the matching tiles. In order to ensure that

the sampled indices from the Level 2A image

products were not affected by clouds, further

processing of the Level 2A image scenes using

cloud masks application, as a byproduct of sentinel-

2 data from the classification, was performed in R.

Several vegetation and biophysical indices including

the NDVI and its derivatives, and the leaf area index

(LAI) were extracted using the created “Snap

Models” function in R-Software environment. The

output products (images), were exported as image

files (Geotiff) into a GIS (QGIS/ArcGIS) software

for visualizations and further analyses, including

overlaying with the plots (sampled sites of the KiLi-

project) for final products output and sharing with

park staff as a web-service maps through the

Tanzania National Parks (TANAPA) Server, located

at the headquarters office in Arusha.

3 ANALYSES AND RESULTS

Sentinel-2 data used here, provides the ability to

carry out consistent monitoring work at high spatial

resolutions ranging from 10m, while providing a

wide opportunity for the use of the Red-Edge bands

(bands 5, 6 & 7) for a vast of biodiversity

measurements for ecosystem monitoring (Rocchini,

et al. 2015). This is because RS allows

measurements of large regions in a short period of

time thereby providing continuous information about

vegetation status. The reflectance and emission of

light from the Earth‘s surface can be directly related

to physiological, morphological and structural

composition of plants (Jetz et al., 2016).

Several studies have proven a significant

correlation between species richness and spectral

indices (Peters, et al. 2016). The most common used

indices are NDVI, capturing the greenness and

chlorophyll content. The green normalized

difference vegetation index (GNDVI), a modified

form of NDVI (Clevers, et al. 2002), which linearly

correlates with LAI, the transformed soil adjusted

index (TSAVI), which corrects the variations of soil

background (Huete, 1988) and a simple ratio

between Red and NIR, RVI; have also been used in

various studies for assessing and monitoring

biodiversity in tropical forest ecosystems (Rocchini,

et al. 2015). All these indices were explored in the

study and proved to be useful in monitoring the key

biodiversity parameters at Kilimanjaro and its

ecosystems, where frequent in-situ observations by

field staff may be difficult given the nature of the

terrain of this unique cloud montane forest.

Several vegetation and biophysical indices were

derived from the final Sentinel- 2 images that were

chosen based on the quality of their scenes output

from the processes described in 2 above, with their

descriptions in Table 2, and a few of the selected

indices are presented in Figure 5.

3.1 Arvi Maps (2019, 2018 and 2017)

The Atmospherically Resistant vegetation index

(Arvi), resistant to the atmospheric effects (in

comparison to the NDVI) is accomplished by a self-

correction process for the atmospheric effects on the

red channel. Arvi takes advantages of the different

scattering responses from the blue and red bands to

Use of Current Remote Sensing Methods for Biodiversity Monitoring and Conservation of Mount Kilimanjaro National Park Ecosystems

179

Table 2: Vegetation indices derived from Sentinel 2-MSI sensor products (images) used in this study between 2017 and

2019, for both tiles T37-MBS and T37-MCS of the Kilimanjaro Mountain ecosystems.

Index Index Application Derived formula from image bands

NDVI

Normalized Difference Vegetation Index (The most

commonly used in RS studies) its values range between -1

and +1 (Rouse, et al. 1973)

NIR-RED

NIR+RED

RVI

Ratio Vegetation Index, also known as Simple Ratio has

high reflectance for vegetation than soil, water and snow-

preferred for mapping vegetation (Jordan, 1969)

NIR

RED

ARVI

Atmospherically Resistant Vegetation Index, accomplishes

self-atmospheric corrections in the red channel (Kaufman &

Tanre, 1992)

NIR-RED-y(RED-BLUE)

NIR+RED-y(RED-BLUE)

TNDVI

Transformed NDVI- preferred because it’s the square root of

NDVI and so its values are always positive (can be larger

than 1) (Senseman, et al. 1996)

√ (NIR-RED + 0.5)

NIR+RED

IRECI

Inverted Red Edge Chlorophyll Index utilizes the Red-Edge

bands (bands 5, 6 & 7) currently present in sentinel 2 data

(Clevers, et al. 2002)

(NIR *NIR-RED1*RED1)

(RED2*RED2/RED3*RED3)

GNDVI

Green Normalized Difference Vegetation Index is strongly

correlated to the leaf area index and hence chlorophyll

content (Gitelson, et al. 1996)

NIR- GREEN

NIR+GREEN

NDI45

Normalized Difference Index is more linearly, less saturated

at higher values than NDVI, (Delegido, et al. 2011b)

(NIR*NIR- RED*RED)

(NIR*NIR+RED*RED)

TSAVI

Transformed Soil Adjusted Vegetation Index is used to

correct the effects of soil line arbitrary values due to

slope/terrain of the area on vegetation (Baret & Guyot, 1991)

(NIR*NIR-s *RED*RED-a)

(a*NIR*NIR+RED*RED-a)

retrieve information regarding the atmospheric

opacity (the blue sky). These properties therefore

have determined that Arvi, has a similar dynamic

range to the NDVI, but is on average four times less

sensitive to the atmospheric effects than the NDVI

(Jetz, et al., 2012). Areas of dense green vegetation

(the montane forest belt) showed clearly high

reflectance (more brightness) compared to non-

vegetated areas of the Kilimanjaro, as shown in

Figure 5.

3.2 Ireci Maps (2019, 2018 and 2017)

The Inverted Red Edge index (Ireci) algorithm

incorporates the reflectance in four bands to estimate

canopy chlorophyll content. The “red edge” is the

name given to the abrupt reflectance change in the

680±740 nm region of vegetation spectra that is

caused by the combined effects of strong chlorophyll

absorption and leaf internal scattering. The position

of the red edge has been used as an indication of

stress and scene sense of vegetation, (Clevers, et al.,

2002). The time series images show darker tones in

areas void of vegetation compared with areas of the

montane forest with brighter tones, for all the years;

2017 to 2019 (Figure 5).

3.3 Tndvi Maps (2019, 2018 and 2017)

The transformed normalized difference vegetation

index (Tndvi) algorithm indicates a relation between

the amount of green biomass that is found in a pixel

and it’s the square root of NDVI. It is superior to

NDVI in that it has higher coefficient of

determination for the same variable and always has

positive values and the variances of the ratio are

proportional to the mean values. Due to limitations

from effects of clouds for obtaining more quality

data, our results indicate that Tndvi could be better

in determining changes of time series related to

monitoring tropical cloud forests, but additional

analyses are needed in the future (seeFigure5).

GISTAM 2020 - 6th International Conference on Geographical Information Systems Theory, Applications and Management

180

Figure 5: Derived vegetation indices from the sentinel-2 data for the Kilimanjaro between 2017 and 2019. Bright areas

represent high values and dark areas low values of vegetation coverage. (Arvi= atmospherically resistant vegetation index;

Ireci=Inverted red-edge chlorophyll index; Tndvi= Transformed normalize difference vegetation index).

4 DISCUSSION AND

CONCLUSIONS

This study attempts to show the use of current

remote sensing sensors capabilities, in particular the

Sentinel- 2 MSI sensor of ESA, in providing data for

ecological, biodiversity and conservation

monitoring, much needed in managing vast

protected areas such as Kilimanjaro Mountain and

its ecosystems. Previous research work indicated

that future climatic characteristics of the Kilimanjaro

ecosystems are mainly determined by the local land-

use and global climate change (Thompson, et al.

2002; Detsch, et al. 2017; Hemp & Hemp, 2018),

and therefore it is imperative that Kilimanjaro

Mountain National Park management is able to

consistently carry out a monitoring program for the

key biodiversity and environmental parameters in

for both current and long-term plans. A workflow

(protocol) developed here, (Figure 4) which is

automated has been customized in such a way that it

can be easily followed up, even with non-RS

specialists, which is usually the case in many

protected areas. Just with a minimum of a computer

that is connected to the ArcGIS Server at TANAPA

headquarters office in Arusha, through the internal

network, the data model workflows output is

uploaded, ran and results can be accessed for sharing

using the available web-services in the ArcGIS

Server network, through the publishing services

(ESRI, ArcGIS Vers. 10.6) .

Results from the derived indices ascertain that it

is possible to monitor key biodiversity changes using

the workflows data-protocols developed here, given

the fact that Sentinel-2 data are also available for

free downloads through the ESA website, and

expected to continue its operations for free data

availability at least until 2028 (Skidmore, et al.

2015). The time-series images indicated that this

technique can be used to show areas of change

(increase or decrease in cover), either through

anthropogenic disturbances such as land-cover

conversions for cultivation and other illegal

activities in the forest and/ or natural phenomenon

associated with climate change, such as recurring

wild fires in the montane forest in previous decades

(Hemp, 2009; 2005). The outputs derived from the

data workflows model will provide early warnings

on the environmental conditions and lead in carrying

out more detailed ground surveys, at focused areas.

Such surveys will guide management decisions for

quick interventions using the established

protocols/methodologies while ensuring cost-

effective park operation undertakings.

The use of the Sentinel-2 data products studied

here will enable an integrated ecosystem

Use of Current Remote Sensing Methods for Biodiversity Monitoring and Conservation of Mount Kilimanjaro National Park Ecosystems

181

measurement and monitoring, through the derived

indices of vegetation like the NDVI- and its

derivatives including TNDVI, ARVI, IRECI,

TSAVI and more as well as the biophysical indices,

such as LAI and the fraction of vegetation cover

(FVC), (Wang, et al. 2018). These indices provide a

bird’s eye view snapshot urgently needed to monitor

these vast areas, while contributing to the global

biodiversity conservation and monitoring agenda,

especially needed in achieving the Aichi

Conservation Targets (2011-2020), in developing

essential biodiversity variables (EBV) from RS data

(Alleaume, et al., 2018; Skidmore, et al., 2015).

Different spectral bands combination derived from

Sentinel-2 sensors, ranging from the Visible, Red-

Edge, Near and Short Infra-red spectra, important

for biodiversity monitoring will provide data from

consistent images of time series indices like NDVI

and its derivatives for rigorous analyses in

biodiversity monitoring of the park and its

surrounding ecosystems.

In order to overcome limitations from accessing

enough continuous image data from scenes that are

cloud free, caused by the nature of the cloud forests,

like in the Kilimanjaro Mountain ecosystems, the

developed data-model protocols provide for

automation of step by step in the selection of

available scenes and enhancement techniques

needed to obtain the final products for further

analyses. Further work in this study would explore

the variances in the ratios obtained for the different

indices’ derived here in relation to each of the

research plots/sampled sites, along the elevation and

across the different land cover/use types gradients in

the study area.

ACKNOWLEDGEMENTS

This work was supported by the German Research

Foundation (DFG), through the KiLi1-Project

(2010-2018), as part of the post-project synthesis

phase for monitoring key biodiversity aspects in the

Kilimanjaro Mountain Ecosystems. F. Msoffe’s time

at Marburg was supported through the Katholischer

Akademischer Auslander Dienst (KAAD)-

Stipendiatum (Scholarship) between 2018 and 2020

as a Post-doc Researcher at the department of

Physical Geography, Umwelti-informatik, Philipps

University, Marburg, Germany.

REFERENCES

Alleaume, S., Dusseux, P., Thierion, V., Commagnac, L.,

Laventure, S., Lang, M., Féret, J-B., Hubert-Moy, L.,

Luque, S., 2018. A generic remote sensing approach

to derive operational essential biodiversity variables

(EBVs) for conservation planning. Methods Ecol.Evol.

2018; 9:1822–1836

Appelhans, T., Mwangomo, E., Otte, I., Detsch, F., Nauss,

T. & Hemp, A., 2016. Eco-meteorological

characteristics of the southern slopes of Kilimanjaro,

Tanzania. International Journal of Climatology, 36,

3245–3258.

Baret, F. and Guyot, G., 1991. Potentials and limits of

vegetation indices for LAI and PAR assessment

Remote Sens. Environ. 35 (1991), pp. 161-173

Burgess, N., Malugu, I., Sumbi, P., Kashindye, A., Kijazi,

A., Tabor, K., Mbilinyi, B., Kashaigili, J., Maxwell,

T., Roy, E.W., Coad, G.L., Carr, K.K.J., Ahrends, A.,

and Newham, R.L., 2017. Two decades of change in

state, pressure and conservation responses in the

coastal forest biodiversity hotspots of Tanzania. Oryx,

2017, 51(1), 77–86

Clevers, J.G., DeJong, S.M., Epema, G.F., VanderMeer,

F.D., Bakker, W.H., Skidmore, A.K.H., 2002.

Derivation of the red-edge index using the MERIS

standard band setting. Int.J.Rem.Sens. 23: 3134-3184

Delegido, J., Verrelsti, J., Alonso, I., Moreno, J., 2011b.

Evaluation of Sentinel-2 red-edge bands for empirical

estimation of green LAI and chlorophyll content.

Sensors 11, 7063-7081.

Detsch, F., Otte, I., Applhans, T. and Nauss, T., 2017. A

glimpse at short-term controls of evapo-transpiration

along the southern slopes of Kilimanjaro. Environ.

Monit. Asses. 189: 465

Drusch, M., Bello, U.D., Carlier, S., Colin, O., Fernandez,

V., Gascon, F., Hoersch, B., Isola, C., Laberinti, P.,

Martimort, P., Meygret, A., Spoto, F., Sy, O.,

Marchese, F., Bargellini., P., 2012. Sentinel-2: ESA's

Optical High-Resolution Mission for GMES

Operational Services. Remote Sensing of Environment,

120: 25–36.

ESA-Corpenicuswebsite-https://sentinels.copernicus.eu/

web/sentinel/sentinel-data-access/registration. Last

accessed September, 2019.

Environmental Systems Research Institute (ESRI),

ArcGIS Software, Version 10.6 (2019), Redlands,

California, USA.

Frampton, W.J., Dash, J., Watmaugh, G. and Milton, E.J.,

2013. Evaluating the capabilities of Sentinel-2 for

quantitative estimation of biophysical variables in

vegetation. ISPRS J. of Photogrametry and Remote

Sens. 82 (2013), 83-92

Gitelson, A.A., Kaufman, Y.J., Merzlyak, M.N., 1996.

Use of a green channel in remote Sensing of global

vegetation from EOS-MODIS. Remote Sens. Environ.

58, 289-298.

Hemp, A. 2005. Climate change-driven forest fires

marginalize the impact of ice cap wasting on

Kilimanjaro. Glob Change Biol. 11: 1013-2013

GISTAM 2020 - 6th International Conference on Geographical Information Systems Theory, Applications and Management

182

Hemp, A. 2009. Climate change and its impacts on forests

of Kilimanjaro. Afr. J. Ecol., 47 (Suppl. 1), 3–10

Hemp, A., and Hemp, C., 2018. Broken bridges: The

isolation of Kilimanjaro's ecosystem: Glob Change

Biol.24:3499–3507.

Huete, A.R., 1988. A Soil-Adjusted Vegetation Index

(SAVI). Remote Sensing of Environment. 25, 295-

309.

Jetz, W., Cavender-Bares, J., Pavlick, R., Schimel, D.,

Davis, F.W., Asner, G.P., Guralnick, R., Kattge, J.,

Latimer, A.M., Moorcroft, P., Schaepman, M.E.,

Schildhauer, M.P., Schneider, F.D., Schrodt, F., Stahl,

U., Ustin, S.L., 2016. Monitoring plant functional

diversity from space. Nature Plants. 2(3), 1-13.

Jordan, C.F., 1969. Derivation of leaf-area index from

quality of light on the forest floor. Ecology 50, 663-

666.

Kaufman, Y.J., Tanre, D., 1992. Atmospherically resistant

vegetation index for EOS-MODIS. IEEE Trans.

Geosci. Remote Sens. 30, 261-270.

Peters, M., Hemp, A., Appelhans, T., Behler, C., Classen,

A., Detsch, F., Ensslin, A., Ferger, S., Frederiksen, S.,

Gebert, F., Haas, M., Helbig-Bonitz, M., Hemp, C.,

Kindeketa, W., Mwangomo, E., Ngereza, C., Otte, I.,

Röder, J., Rutten, G., Schellenberger Costa, D.,

Tardanico, J., Zancolli, G., Deckert, J., Eardley, C.,

Peters, R., Rödel, M., Scheleuning, M., Ssymank, A.,

Kakengi, V., Zhang, J., Böhning-Gaese, K., Brandl,

R., Kalko, E., Kleyer, M., Nauß, T., Tschapka, M.,

Fischer, M., Steffan-Dewenter, I., 2016. Predictors of

elevational biodiversity gradients change from single

taxa to the multi-taxa community level. Nature

Communications 7: 13736

Rocchini, D., Boyd, D., Feret, J.-B., Foody, G., He, K.,

Lausch, A., Nagendra, H., Wegmann, M., Pettorelli,

N., 2015. Satellite remote sensing to monitor species

diversity: potential and pitfalls. Rem. Sens.Ecol and

Conserv, Zoological Society of London (2015),

Review. 25-36.

Rouse, J.W., Haas, R.H., Schell, J.A., Deering, W.D.,

1973. Monitoring vegetation systems in the Great

Plains with ERTS, In: Third ERTS Symposium,

NASA SP-351, pp. 309-317.

Senseman, G.M., Tweddale, S. A., Anderson, A. B. and

Bagley, C. F., 1996. Correlation of Land Condition

Trend Analysis (LCTA) Rangeland Cover Measures to

Satellite Imagery-Derived Vegetation Indices-

http://www.cecer.army.mil/techreports/and_vegi/AND

_V EGI.LLN.post.PDF

Tabor, K., Hewson, J., Tien, H., Gonzales-Roglich, M.,

Hole, D., and Williams, J.W., 2018. Tropical Protected

Areas Under Increasing Threats from Climate Change

and Deforestation. Land 2018, 7, 90;

Doi:10.3390/land7030090

Thompson, L.G., Mosley-Thompson, E., Davis, M.E. et

al., 2002. Kilimanjaro ice core records: evidence of

holocence climate change in tropical Africa.

Science298, 589–593.

Skidmore, A. K., Pettorelli, N., Coops, N. C., Geller, G.

N., Hansen, M., Lucas, R., Wegman, M., 2015.

Environmental Science: Agree on biodiversity metrics

to track from space. Nature, 523, 403–405.

Wilm, U., 2018. S2 MPC. Sen2Cor Configuration and

User Manual. Ref. S2-PDGS-MPC-L2A-SUM-V2.2.5.

Use of Current Remote Sensing Methods for Biodiversity Monitoring and Conservation of Mount Kilimanjaro National Park Ecosystems

183