Schedule

Subscribe to RSS Feed

2020
Tuesday, September 15th
4:40 PM

Assessing Urban Vulnerability and Resilience to Extreme Heat with Big Data and Machine Learning: The Case of July 2019 European Heatwave

Mikhail Sirenko, Delft University of Technology

4:40 PM - 5:00 PM

Climate change has put residents of many cities around the world at risk. The challenge for urban planners is not only to understand which people are vulnerable and where they are located but also assess how resilient the city is to climatic shocks. To assess vulnerability and resilience, we proposed a framework based on open data and open-source machine learning libraries. We applied this framework to analyze the impact of July 2019 European heatwave on The Hague, the Netherlands. The framework combines a geodemographic grid of 500 by 500 m2 with 3,973,549 anonymized ambulance calls to capture citizens' vulnerability and street networks with distance matrices for evaluation of resilience as accessibility by critical infrastructure. We found that vulnerability and resilience are unequally distributed both spatially and temporally. People who are the most sensitive to extreme heat such as low-income single households with kids and elderly with mobility issues are also exposed the most. The lack of greenery in their residential areas resulted in urban heat island effect which may have amplified the impact of the heatwave. Such a combination represents high vulnerability and has led to an increase in ambulance calls by 17% from the yearly average. The south and northwest parts of the city where the population is dominated by above mentioned groups by 75% are the furthest from the hospitals and as a result less resilient. In addition to an increase in calls, new peak hours appeared at 12:00, 16:00 and 02:00 where intense afternoon and evening traffic make these citizens hardly reachable when they are the most vulnerable. Knowledge of spatial and temporal variability in vulnerability and resilience can provide policy-makers with insight about potential interventions. Our proposed framework can be generalized and applied to other cities with similar data availability.

5:40 PM

Use of Remotely Sensed Data in Modelling Environmental Quality: Case Study of Selected Kenyan Groundwater Basins

Silvance Abeka, Jaramogi Oginga Odinga Univers

5:40 PM - 6:00 PM

Kenya is classified as a water scarce country with available annual per capita water of less than 1000m3. Quality of environment has drastically deteriorated in the recent past as population increases and by extension increase of unsustainable economic developmental activities. One phase of environment which has suffered quality deterioration in Kenya is water. Groundwater sources which majorly support water needs in arid and semi-arid zones of Kenya which is about seventy five percent of the country is now threatened by pollution. Interactions of surface, subsurface systems and human systems require that integrated studies become the preferable way forward for studying groundwater vulnerability mapping. This study has reviewed effectiveness of GIS-based environmental models in mapping groundwater quality of basins along the Kenyan coast which suffer from seawater encroachment and hinterland basins which suffer salt intrusion geologic unconformities. Softwares as GALDIT and DRASTIC are applied to assist in overlaying relevant thematic maps. Overlaying of carefully selected thematic digital maps derived from remotely sensed information was done for selected basins to produce vulnerability maps which can provide useful insight or effective groundwater management of the study areas.

6:20 PM

Software for Monitoring Forest Change in Tropical West Africa Using Satellite Remote Sensing

Michael Wimberly, University of Oklahoma, USA

6:20 PM - 6:40 PM

Satellite remote sensing is an essential source of data on forest landscape change that is necessary for estimating carbon stocks, assessing biodiversity, and developing conservation plans to sustain ecosystem services. The Upper Guinean Forest region of West Africa, a global biodiversity hotspot, currently lacks timely, consistent, and accurate long-term forest monitoring systems suitable for decision support. Major limitations include data gaps due to frequent cloud cover and limited satellite coverage. To address these limitations, we developed the West African Forest Degradation Data System (WAForDD) to use all available Landsat imagery to monitor forest change in the tropical forest zone of Ghana. The approach used a novel combination of spectral mixture analysis (SMA), machine learning classification, and time-series modelling implemented on the Google Earth Engine (GEE) cloud computing platform. Outputs included annual maps of canopy cover for 2000-2019 at 30 m resolution along with annual change estimates of forest loss, degradation, and recovery. We found that random forest predictions of canopy cover based on SMA fractions resulted in more accurate predictions than those based on other types of spectral indices. Postprocessing the canopy cover maps with the LandTrendr algorithm for segmented time series regression further improved map accuracy. The final maps had an observed-predicted correlation of 0.87 and a mean absolute error of 12.7%. Examination of change trajectories found that decreases in canopy cover coincided with known historical disturbances such as wildfires, logging, land clearing for agriculture, and mining activities. By leveraging the GEE platform and co-developing the software with partners in Ghana, we were able to help the Forestry Commission of Ghana meet their measurement, reporting, and verification (MRV) needs for REDD+ implementation and ensure that use of the software can be sustained into the future.

6:40 PM

Automated Bathymetric Mapping Using Unmanned Aerial Vehicles for Wetlands Modelling

Elizabeth Basha, University of the Pacific, United States

6:40 PM - 7:00 PM

Understanding and managing wetlands requires modeling and monitoring their environments. Wetlands models are improved with more detailed bathymetric maps, which are developed through monitoring. This paper demonstrates how to achieve this monitoring through the use of Unmanned Aerial Vehicles (UAVs). UAVs provide a monitoring method with limited impact on the ecosystem, the option to dynamically balance spatial and temporal monitoring, and the possibility of autonomous monitoring. First, this paper develops, implements, and analyzes the various bathymetric mapping options for UAVs, including point selection and interpolation phases. Through the analysis and simulation results, it determines the best approach to take with an unknown environment. Second, to perform the measurements, a system is developed to measure water depth using a UAV that consists of a depth sensor payload and winch system. The system is incorporated onto a commercial UAV, controlling its flight and verifying its use to create autonomous bathymetric maps.

7:00 PM

Integrated workflow for use of unmanned aerial vehicles in monitoring water resources

Mary Kay Camarillo, University of the Pacific, United States

7:00 PM - 7:20 PM

Unmanned aerial vehicles (UAVs) offer possibilities to enhance monitoring of river basins, wetlands, and other important water bodies. Opportunities to use UAVs in environmental studies include periodic and autonomous flights to detect and provide warnings of algal blooms, access to areas that are difficult to explore, and more rich data sets of water coverage, vegetation, and other features. In order to benefit from UAV technology, appropriate flight planning, data collection, and post-flight data processing strategies must be developed. In this study, we present recommendations for integrating UAVs into the study of water bodies. We first provide an overview of the key aspects of flight planning that take into account the desired accuracy and number of images required. Given the computational resources required to generate aerial maps, we consider merging of adjacent point clouds. We evaluate ground control point placement to connect maps into existing coordinate systems and improve map accuracy. The workflow is demonstrated using a river basin located in an urban setting that is bound on both sides by flood-control levees. A 2.7 ha area was mapped using 200 images with a ground sampling distance of approximately 1.65 cm/pixel. Average root mean square error along the levees was 3.1cm in the horizontal plane, and 6.9cm in the vertical direction. Error near the river was much higher (35-40cm), suggesting more control points be placed near water bodies. The results of this study demonstrate that UAVs can be introduced into water resource monitoring, but that UAV integration into current projects requires planning, resource investment, and education on the part of the users of such technology.

Thursday, September 17th
3:20 PM

Cyanobacteria Assessment Network Cross-Platform Responsive Web Application (CyANWeb)

John Johnston, EPA/ORD, USA, Environmental Processes Division, Athens, GA, USA

3:20 PM - 3:40 PM

The Cyanobacteria Assessment Network (CyAN) is a multi-agency (NASA, USGS, NOAA, and EPA) project whose mission supports the environmental management and safe public use of U.S. lakes and reservoirs by providing a capability of detecting and quantifying algal blooms using satellite data records and disseminating this information through a mobile application developed and hosted by EPA. The Cyanobacterial Index (CI, a measure of abundance) is derived from the European Space Agency’s Copernicus Sentinel-3 satellite Ocean and Land Colour Instruments (OLCI) and covers the continental U.S. at 300m resolution and represents the weekly maximum value of cyanobacteria at over 2,370 resolvable lakes and reservoirs. The imagery is updated weekly, with previous week’s data available mid-current week, to provide a near-real time view of cyanoblooms at user selected areas of interest. The CyAN Android mobile app has been publicly available since June 2019 with a reported positive user experience (4.0/5.0 rating and 1000+ installs); however, the mobile app is Android-device operationally only. To reach a broader userbase and increase CyAN app flexibility, the EPA has undertaken development of a device-agnostic responsive web application (CyANWeb) usable on most internet-connected devices via a web browser (and operational on any device operating system platform). CyANWeb uses the JavaScript Angular development framework to achieve its cross-platform dynamic website design goal. CyANWeb is a new client application that mirrors the functionality of the Android mobile app and serves the current OLCI CI on an ongoing basis. Future enhancements include aggregate cyanobloom statistics by waterbody, weekly reports, and user-requested location CI data to be implemented in CyANWeb. CyANWeb will begin beta-testing in the summer of 2020 and run in parallel with the Android mobile app.

Optimizing Sustainable Seafood Production through Predictive Machine Learning Algorithm

Peter Khaiter, York University

3:20 PM - 3:40 PM

Abstract missing