How to find relevant climate data

A new tool helps users navigate the hundreds of open climate data platforms available and identify the most relevant platforms for their work.

The 2015 Paris Agreement outlines the mitigation, adaptation, and financial transitions required to respond to current and future climate change impacts and champions a just transition to help the communities most affected. The systemic transitions needed for a climate-safe future will require many types of data — including at both the country and city levels, across both adaptation and mitigation measures, and efforts to address both public officials and the private sector. 

The good news is that many aspects of climate change are measurable, and this data can help identify problems, inform decisions, and set up timely information systems. Data platforms — websites that let users view, explore, and download a variety of datasets — have emerged to support the curation and sharing of this growing ecosystem of climate data. However, with hundreds of platforms available, it can be difficult to find the best data for your particular needs or to know where data gaps still exist. To remedy this challenge, we created a map of open climate data platforms called the Climate Data Platforms Explorer to help users find the most relevant platforms for their work, understand which initiatives are underway, and identify where new research can add the most value.

However, with hundreds of platforms available, it can be difficult to find the best data for your particular needs or to know where data gaps still exist. To remedy this challenge, we created a map of open climate data platforms called the Climate Data Platforms Explorer to help users find the most relevant platforms for their work, understand which initiatives are underway, and identify where new research can add the most value.

The explorer shows a matrix of more than 80 major data platforms identified through this process1, displayed by topic (x-axis) and geographic level (y-axis). Tools that cover several topics or scales are shown multiple times. Platforms were selected for inclusion based on a set of criteria including the relevance of data for climate change science, policy, and action; a relatively broad geographic scope; reasonably up-to-date data; and robust methods.

The Climate Data Platforms Explorer allows different users to discover the most relevant data for their work using filters in a visual format, a filterable table, or a download of the raw data. In curating this dataset of major platforms, we identified several notable patterns in the climate data ecosystem. 

Most data platforms focus on mitigation, energy, and country-level data

A majority of data platforms focus on climate change mitigation — addressing greenhouse gas emissions and the economic activities that cause them — and on the energy sector. Both areas are core concerns: addressing climate change requires rapid decarbonization, and the energy sector is responsible for 76% of global emissions. Similarly, there is greater coverage at the country and regional levels than at more granular geographic levels. 

There are significant gaps in data and platforms around topics that are harder to measure like equity, policy, adaptation, and resilience, and at the subnational level. Accessible data is needed to support climate change across sectors and different actors, including through city, corporate, and local climate action. Adaptation efforts, in particular, require local data that is often unavailable or insufficient. 

A new platform isn’t always the solution

We identified many similar platforms that showcase relatively similar datasets. It is not always clear that these datasets needed to be hosted on an entirely new platform, and they may have been more effective if integrated into existing resources. 

Any new platforms — and new datasets powering these platforms — should ensure that they build on existing efforts, align with previous standards, and show that they provide actionable data. In considering new work, project leads should first engage in robust scoping to understand what the most urgent and actionable data gaps are, and how new datasets or new data platforms might or might not fill these gaps. This should include interviews with both supporters and skeptics, as well as user testing to understand what target audiences need from a platform and whether the planned platform would be able to meet these needs. Second, we also suggest exploring simpler ways to make data available to potential users before building out a complex website — such as a simple provision of the raw data with summary tables and a brief methodological explainer. Finally, while the transaction costs of collaboration can be challenging, we encourage practitioners to consider whether they can meet their aims by incorporating their datasets into existing platforms rather than building out new tools.

Data maintenance plans are sometimes an afterthought

Some of the platforms that we assessed for this project appeared to be carefully built, but are no longer maintained. We encountered a variety of issues, including server problems, broken data links, visuals that relied on outdated software, or outdated data. In some cases, there may be good reasons for these issues — for example, a platform may have relied on a dataset that was retired, or the research was time-bound. However, in other cases, the problems were tied to age-old problems, such as lapsed funding, staff turnover, or a lack of capacity for maintenance. 

If a new platform appears warranted, its maintenance should be a prime consideration. Building an entirely new data platform has costs, and limited research funds could be put to more urgent use than this creation-and-abandonment cycle. It is essential for researchers to build long-term maintenance plans and for funders to be clear about the potential for longer-term support or lack thereof. 

Building a better climate data system

It is clear that we need to improve our climate-related data infrastructure. As suggested earlier, the requisite data often exist, but they are collected by different types of users for different purposes, and scattered across multiple platforms with varying standards. Consequently, it is difficult to discover and use this data for real-world applications.

To address these problems, we need to articulate a vision of a system of open, shared, and integrated data. Ideally, existing climate data would be open by default, discoverable, and digitally accessible; data standards to promote interoperability, efficiency, and user flexibility would evolve in response to user demand; and researchers, funders, governments, and NGOs would fill data gaps by sharing high-quality data, along with essential metadata. These actions could lead to an interconnected system of public-facing climate data to meet the diverse needs of data users. The first step in creating a connected climate data ecosystem is identifying credible climate-relevant data platforms around the world.

The Climate Data Platforms Explorer offers first steps in this direction by curating a set of climate data platforms to support users in finding the most relevant platforms for their work, providing a picture of what work is already underway, and identifying where new research can add the most value. 

A call to action for researchers, funders, and governments

Researchers, philanthropic funders, and governments can take the following steps to move toward a more connected and rationalized climate data ecosystem. If you are considering any new platforms, we suggest that your work should: 

• Avoid redundancies and confirm user needs: Conduct due diligence to ensure that new datasets and platforms are complementary to and not redundant with existing efforts — and that they address confirmed user needs. Where possible, build with or on existing platforms and standards rather than launching new projects. Those creating or commissioning new data should consider earmarking support within project budgets to enable this due diligence, user testing, and the increased transaction costs that collaboration can require.
• Be open, interoperable, and discoverable: If new datasets and/or platforms prove necessary, they should be designed to be open and interoperable, allowing data and metadata to be shared within existing repositories. Good metadata could enable a decentralized interconnected system — similar, for example, to Data.gov, in which data creators could submit their data and platforms to a central inventory. 
• Include a plan for long-term maintenance: Finally, it is critical to plan for the maintenance of the dataset and the platform on which it lives over time. Questions to consider include: is the platform tied to a time-bound event or will it exist in perpetuity? Is there a plan to sunset it if new data is not forthcoming? Who on the team owns ongoing maintenance tasks, and is there a budget earmarked for repeating costs such as hosting services? Funders can support these efforts by surfacing these questions before supporting new work and building longer-term support into project budgets.