Scanning for Lifeforms
Visualizing the evolution of artificial night light encroachment on natural protected areas
Summary
Due to increased urbanization, the proportion of dark areas within protected sites has been decreasing. Because light pollution represents a barrier to species movement between otherwise adequate habitat patches (among other negative impacts), protected areas would benefit from incorporating this issue into their management strategies. We propose that an effective visualization of the progressive encroachment of light pollution on protected areas could increase awareness of the issue, and provide timely evidence to promote its inclusion as part of the criteria to define and manage these sites of paramount importance to biological diversity.
How We Addressed This Challenge
Protected areas are a conservation tool of paramount importance in light of the accelerated land-use change caused by human activities. By covering large extensions of habitat, they preserve not only the species than inhabit them, but the ecological processes and biological interactions that are critical to ecosystem functioning, and that provide humans with a wealth of ecosystem services. While the intention of protected areas is to ensure that their designation buffers them against severe impacts of human activities, this is not always the case. In particular, sources of diffuse pollution can infiltrate and cause negative effects on the ecological integrity of these areas. One such source is light pollution. The growth and urbanization of human settlements has been accompanied by an increased use of electric lighting. Although of importance to enhance the quality and productivity of human activities, particularly at night, widespread use of artificial light has transformed the nocturnal landscape. Calculations estimate an average 6% annual increase in artificial light at night (ALAN) emission, with over 23% of the land surface of the planet experiencing increased levels of night sky brightness due to skyglow, a consequence of light scattering and rebounding on atmospheric particles that creates a halo of light that can affect areas many kilometers away from the source point.
We developed an online platform where the user could interact with data layers that would enable them to visualize how light pollution has encroached on protected areas over time. We propose that an effective visualization of this issue lies at the core of raising awareness of a problem that is otherwise not very well known outside the academia. While it is true that there are movements that advocate for reducing light pollution (such as the dark skies movement), it is less associated in the public's mind to negative effects on biodiversity, as demonstrated by the fact that light pollution is seldom considered as a criteria for defining or managing protected areas.
The platform we developed currently shows changes in light pollution between 1992-2008 in 5 countries (Laos, Thailand, Cambodia, South Africa and the Republic of Korea), selected on the basis of the relative increase of artificial light at night and their high biodiversity. The map displays annually averaged NTL measurements in a global grid with a 30 arc-seconds (~1 km) spatial resolution, and its overlap with terrestrial protected areas of those countries. A drop-down menu allows the user to select the year of display.
While still in development phase, we hope that visualizing both datasets together will provide timely evidence to prompt the inclusion of light pollution as part of the criteria to define and manage protected areas.
How We Developed This Project
One of the crucial challenges to preserve biodiversity is to understand how human activity drives changes in the environment that in turn affect ecosystem function, animal behaviour and ecosystem services (among others). Some of these changes are evident in terms of their impacts (for instance, land use change) and are very present in the public eye. However, other human-driven effects on the environment are often overlooked, with unintended and unattended consequences. We chose to focus on light pollution as an example of how remote sensing can help us see the forest from the trees, both in a figurative and literal sense, and raise awareness on an issue which had been overlooked until recently, despite its major impact across many levels of natural ecosystems.
The project is based on the analysis of three types of datasets: global satellite maps of nighttime light (NTL), geographic polygons delimiting natural Protected Areas (PAs), and geolocated fauna sighting reports. After download from the public repository, the NTL data was processed with Python using the rasterio geospatial data processing module to (1) convert it into a JSON format compatible with the deck.gl HeatmapLayer specification, and (2) down-scale the global geographic data into 10 km and 100 km grids for faster processing and reduced disk and bandwidth usage, albeit selected geographic areas were extracted at the original 1 km resolution. Analysis of the NTL data and its relation to the Protected Area polygons was done with the help of the rasterstats package. Python3, numpy and matplotlib were used for general data handling and plotting. PAs data were obtained from the World Database on Protected Areas (WDPA), downloaded in a shapefile format and converted to geojson format using the geopandas library in python, selecting only the terrestrial protected area polygons.
Django and GraphQL were used to create an app for data management and visualization, while the integration of Docker made it possible to deploy it in any UNIX environment. The GBIF API was integrated to manage the records of occurrences of different species. The necessary models were developed for future automatization to obtain the data used in this project through the integration of different APIs.
Although the original intention was to incorporate data on species occurrences and movement (specifically bats, using data from Gbif and Movebank), and to increase the resolution of our analysis by including data from the Cities at night project (http://citiesatnight.org/), we ran out of time to search for the specific type of data (longitudinal data) that would have allowed us to track and assess for differences in species distribution.
How We Used Space Agency Data in This Project
The NTL satellite data studied are those published by Li et al. (2020), obtained by the integration and harmonization of satellite data of the Defense Meteorological Satellite Program and of the more recent Day/Night Band of the VIIRS sensor onboard the NOAA-20 and Suomi NPP meteorological satellites. The dataset contains annually averaged NTL measurements in a global grid with a 30 arc-seconds (~1 km) spatial resolution in the 1992-2018 period. Geospatial time series from these satellites were re-calibrated, normalized and integrated into a single data product expressing NTL as discrete 64-level digital values. Sources of noise such as aurora and transient lights were removed from the data. The data is publicly available in a data repository in GEOTIFF format
Project Demo
Project Code
Data & Resources
- Satellite data: Defense Meteorological Satellite Program and of the more recent Day/Night Band of the VIIRS sensor onboard the NOAA-20 and Suomi NPP meteorological satellites
- Terrestrial Protected Areas World Database on Protected Areas (WDPA)
- GBIF data on species occurrences