The assessment of UAV-based multispectral imagery to monitor lichen-dominated biological soil crusts response to fog and disturbances, Central Namib
No Thumbnail Available
Date
2022-10
Authors
Journal Title
Journal ISSN
Volume Title
Publisher
Namibia University of Science and Technology
Abstract
Lichen-dominated biological soil crusts (BSCs or biocrusts) are a major contributor to biodiversity and
ecosystem services of desert environments. There appropriate monitoring of lichen fields of the hyperarid
coastal Namib Desert is important to understanding their greater ecological functions, such as vital
CO2 sinks, and for management of socio-economic land use practices (i.e. mining, tourism, off-road tracks)
which destroy the endemic, fragile and slow recovering lichens. Thanks to advanced remote sensing
techniques such as the use of Unmanned Aerial Vehicles (UAVs or drones), lichen fields can be monitored,
whilst minimising the environmental impact of traditional ground-based ecological research, as well as
overcoming challenges in monitoring lichens with the satellite datasets, piloted aircraft. The objectives of
this study was to assess the effects of hourly sunshine intensity on the use of UAV-based multispectral
imagery for monitoring lichen-dominated biological soil crusts in the hyper-arid Central Namib Desert. We
then assess the applications UAV-based NDVI to observe lichen photosynthetic activity response to
moisture input, topography and disturbances, as well as to assess the influence of micro-topography and
disturbances in off-road tracks on lichen coverage.
We tested the performance of UAV-based Parrot Sequoia multispectral sensor by acquiring time series
flights from 08:00 to 17:00 during a fog day. Time series of radiometrically and atmospherically corrected
and uncorrected multispectral image datasets (i.e. reflectance images and vegetation indices) were
compare it to solar elevation. Influence of brightness effects in sample plot on image values was tested
with one-way ANOVA and Tukey HSD pot-hoc test for pairwise comparisons. There was statistical
significant difference in means of the sample plots with p < 0.05. We also gathered time series UAV-based
NDVI images on fog and non-fog day derived the temporal patterns of lichen photosynthetic activity in
responding to wet and dry conditions, overtime. These UAV observations were supported with ground
cover observation of lichens. Statistical analysis was performed to determine presence of significant
differences of lichen cover and NDVI response between micro-topographic habitats (i.e. ridges, windward
and leeward sides of the relief), and as well in landforms such as plains, river washes and tracks. The oneway
ANOVA and Student-test (t-tests), were considered significant at p < 0.05 for both UAV and ground
15
observations. Polynomial regression analysis of the relationship between elevation and ground lichen
cover and NDVI was significantly correlated with p-values of < 0.05.
Well illuminated good quality images can be captured at low solar elevation, in morning images at 08:00
and in the later afternoon at 17:00. Midday images taken between 10:00 and 15:00 at higher solar
elevation are heavily impacted by brightness effects. The radiometrically corrected images and NDVI were
more accurate than the raw uncorrected images. UAV-based NDVI can estimate photosynthetic activity
when lichens are hydrated by fog or higher air humidity, as opposed to when they are desiccated and
photosynthetic activity is inhibited. Elevation is good predictor lichen cover and photosynthetic activity.
Lichen cover is heavily impacted by off-road tracks. The topographic attribute influence the distribution
of lichen cover and lichen species diversity. This research study has provided insights into a means to
monitor these sensitive environments. It also enables informed decision making to manage natural
resource and regulate socio-economic land-use.
Description
Thesis submitted in partial fulfillment of the requirements for the degree of Master
of Natural Resources Management at the Namibia University of Science and
Technology
Keywords
Unmanned Aerial Vehicle (UAV) remote sensing, biological soil crusts (BSCs), Namib Desert Lichen fields, multi-spectral imagery, fog-driven ecosystems, off-road driving disturbance
Citation
Nambwandja, A.N. (2022). The assessment of UAV-based multispectral imagery to monitor lichen-dominated biological soil crusts response to fog and disturbances, Central Namib [Unpublished Masters Thesis]. Namibia University of Science and Technology.