Describe what your organization does, and how it relates to data science.
We train early and mid-career professionals to use data to address complex environmental issues. The Environmental Observation and Informatics MS program, situated in the Nelson Institute for Environmental Studies, is a professionally-oriented, interdisciplinary, and data-driven Master of Science program. Rooted in the discipline of environmental remote sensing, students receive training in utilizing data from various remote platforms such as satellites, airplanes, and drones, along with other geospatial data and tools.
Remote sensing, also referred to as earth observation, is inherently data-intensive and is motivated by the imperative to generate actionable information. In this field, the core data comprises digital images, each containing hundreds of thousands of pixels providing information about the Earth’s surface location and characteristics. Beyond mere aesthetic appeal, these digital images hold a wealth of information about our environment.
Within the realm of remote sensing, common data tasks involve classification, wherein each pixel is assigned a meaningful category. A notable example is land cover classification, wherein remote sensing data and classification techniques enable the creation of maps depicting the distribution of forests, urban areas, lakes, and fields worldwide over time. Notably, remote sensing data scientists contribute significantly to understanding phenomena such as the rate of deforestation in the Amazon rainforest.
Moreover, advancements in these data applications extend to mapping biodiversity and other critical aspects of our environment. The Environmental Observation and Informatics MS program comprehensively exposes students to both dimensions of this work—1) encompassing the data, technology, and tools, and 2) understanding the interdisciplinary environmental applications and the imperative for information that propels the field forward. This ensures that graduates are well-equipped to tackle intricate environmental challenges through the proficient utilization of data science techniques.
Why is sponsoring the Research Bazaar important to you?
We are sponsoring the Research Bazaar to emphasize the significance of environmental applications in data science. In our perspective, training in data science should not be confined to any specific major or discipline. Instead, we aim to demonstrate that data science is integral to the environmental sector, and UW-Madison stands out as a leader in providing interdisciplinary environmental training. This leadership extends to the adept utilization of diverse data sets, including remote sensing, geospatial, and other environmental data. By supporting the Research Bazaar, we contribute to showcasing the pivotal role of data science in addressing environmental challenges and further establishing UW-Madison’s prominence in fostering interdisciplinary expertise.
Could you highlight any unique or cutting-edge technologies that your organization utilizes within the realm of data science?
Open data and cloud-based analysis platforms have become integral components of our environmental work. In this context, a noteworthy trend involves accessing extensive datasets from both public entities such as NASA and private companies like Planet. These datasets, often comprising terabytes of data, are hosted on servers. Instead of the conventional approach of bringing the data to the code, there is a paradigm shift towards bringing the code to the data. While the use of cloud-based platforms and APIs is not groundbreaking in itself, within the field of remote sensing, it has introduced novel dimensions to data science. This shift has significantly enhanced accessibility to data, rendered workflows more transparent, and facilitated the sharing and evaluation of results. Embracing these technologies, however, necessitates individuals to be proficient not only in environmental disciplines but also in these technological skills. This is precisely the rationale behind the existence of the Environmental Observation and Informatics (EOI) program – to equip professionals with the necessary expertise in both environmental science and cutting-edge technology.
How has the integration of data analytics or artificial intelligence influenced the product or service offerings of your organization?
In our field, we have observed a shift towards employing more sophisticated data analysis methods. While simple linear regressions or clustering methods were previously deemed sufficient, current research highlights the effectiveness of advanced techniques such as convolutional neural networks, instance segmentation pipelines, and spatial statistics. This evolution presents both an opportunity and a challenge in training the next generation of professionals. The opportunity lies in aligning environmental applications with the cutting-edge methods that data scientists are increasingly adopting. This creates a chance to attract individuals interested in the field, ensuring that the training remains at the forefront of contemporary data science practices. However, the challenge arises from the fact that individuals with an environmental background face a steeper learning curve in mastering these new data analytics techniques. Given that the Environmental Observation and Informatics (EOI) program operates on a short 15-month timeline to reintegrate professionals into the workforce, strategic training approaches become essential to effectively equip them with the required skills.
How does your organization balance the need for data-driven decision-making with privacy and security considerations?
Within our organization, the awareness of ethical data use, privacy, data security, and access is deemed crucial for every environmental data scientist. While these aspects may not have been prominently emphasized in early remote sensing training, recent initiatives within the geospatial fields and professional certifications have sought to establish ethical codes of conduct.
In the Environmental Observation and Informatics (EOI) program, we take a proactive approach to addressing these considerations. Students actively engage in scenarios that prompt them to contemplate issues such as the impact of geotagging photos on social media when visiting endangered wildlife, the societal implications of open access to all satellite data, and the unintended consequences of publishing a map highlighting areas of high pollution. Recognizing that professionals in the field will inevitably encounter such ethical dilemmas, we guide our students in establishing robust ethical values for their work. This proactive approach ensures that graduates are not only well-versed in data-driven decision-making but also equipped to navigate the complex landscape of privacy and security considerations in the realm of environmental data science.
Are there any notable collaborations or partnerships your organization has formed to advance data science research or applications?
Our organization is privileged to foster numerous collaborations and partnerships, as each student actively engages with an organization for their final project. Several notable examples include:
- Greenlink Analytics (non-profit in Atlanta): Collaboration involves integrating information about the tree canopy and the extent of impervious surface cover into a scalable model focused on health, housing, and environmental equities.
- NASA DEVELOP (federal training program): Collaborative efforts with NASA DEVELOP extend to modeling wildfire risk in Idaho, providing valuable insights to stakeholders for enhanced management strategies.
- Conservation International (global non-profit conservation organization): Our partnership with Conservation International entails evaluating the effectiveness of conservation programs for indigenous communities. The focus is on assessing the preservation of carbon stored in tropical forests and determining the impact on improving livelihoods.
- Dane County Office of Energy and Climate Change (municipal office in Wisconsin): Collaboration with the Dane County Office involves mapping the urban tree canopy to identify inequities related to socioeconomic variables. This effort aids in pinpointing areas for investment in tree planting, contributing to a more equitable and sustainable urban environment.
These collaborations exemplify our commitment to applying data science in addressing real-world challenges and contribute to meaningful advancements in environmental research and applications.
How does your organization collaborate with environmental agencies, research institutions, or NGOs to provide students with opportunities to engage in meaningful, data-driven environmental projects?
As a Master of Science program specifically tailored for professional training, our approach involves facilitating direct engagement between each student and a partner organization in the execution of a data-driven environmental project. These projects typically span a duration of 8-12 weeks and are meticulously designed to be mutually beneficial. They serve the dual purpose of supporting the organization’s objectives while simultaneously advancing the students’ professional skills by applying their training in a real-world, professional setting.
Our extensive network includes partnerships with organizations across the private sector, non-governmental organizations (NGOs), government agencies, and academic institutions, totaling 40 collaborations globally. These projects yield diverse deliverables, ranging from automated data analyses to story maps enriched with data visualizations and comprehensive scientific reports.
The thematic spectrum of these projects encompasses vital environmental domains such as forest resources, agriculture, environmental justice, recreation, water quality, and biodiversity. This collaborative model ensures that our students gain hands-on experience, contributing to the actualization of impactful, data-driven solutions in diverse environmental contexts around the world.
How does your master’s program prepare students to communicate complex environmental data to diverse audiences, including policymakers, communities, and other stakeholders?
Our master’s program employs a multifaceted approach to equip students with the skills necessary to effectively communicate complex environmental data to diverse audiences. The key component of this training is the practical engagement with clients, either through our client-based courses or in their final independent projects. This immersive experience compels students to comprehend the perspective of their clients and tailor their communication strategies accordingly. Different clients may have distinct needs, ranging from a focus on technical details to a requirement for high-level outputs demonstrating the practical utility of the technology. Recognizing the audience and understanding their information needs emerges as a critical skill in this process.
Furthermore, students leverage a variety of communication platforms to convey their findings. This includes the use of maps, incorporating principles of cartography and design, as well as creating story maps and web-based multimedia visualizations. By diversifying the communication tools, students become adept at presenting data in formats that resonate with different stakeholders.
Complementing these technical aspects, our professional development courses address broader communication skills. Students are coached on delivering effective science presentations and engaging with the media. These courses impart fundamental best practices, ensuring that our graduates possess a well-rounded skill set for effectively conveying complex environmental data to a wide array of audiences.