COS 261-5 - A decade of Daymet data: Open source science leverages standardized interoperable data for all

Abstract: The goal of this study is to analyze a decade of distribution of the Daymet dataset from the NASA ORNL DAAC, summarizing results in how Daymet data are distributed and accessed which has resulted in the Daymet dataset ranking as the10th most downloaded dataset from the NASA Earthdata Data Systems (ESDS) program. We'll analyze the methods in which Daymet data are accessed, summarize research domains from journal articles citing the Daymet data, and demonstrate how community resources pertinant to the ecological field are built from open API's.

We'll demonstrate the importance of data standardization and how providing data retrieval methods, both open and specialized, has enabled greater accessibility and allowed the creation of higher-level domain specific frameworks, many of which are directly applicable to the ecological community. Daymet is distributed in the multidimensional netCDF 4 data format, highly standardized to Climate and Forecast (CF) conventions. This level of standardization allows data to be highly interoperable and programmatically accessible from a number of application programs that can transform, combine, analyze, or display specified fields of the data. Daymet V3 has had over 56,000 direct downloads which does not account for other popular programmatic access methods. The specialized access method called the Single Pixel Extraction Tool allows programmatic access to the data in a simplified csv format. It is not unusual for the Single Pixel Tool to have hundreds of thousand of weekly queries. Custom community applications built on top of these programmatic access methods including HyRiver and a suite developed by BlueGreen Labs.

The results of an analysis of download metrics will examine patterns of how researchers access data and the importance of providing machine interoperable and programmatic services. A summary and analysis of almost 600 dataset citations and metrics will show the diversity of study domains to which these data are applied and provide guidance for tutorial development provided by the ORNL DAAC and other groups such as ESRI.

As NASA Earthdata strategizes to operating in a commercial cloud environment, we'll discuss the considerations of how to continue to provide access for all to this highly downloaded and ecologically relevant dataset. Daymet is a 40+-year meteorological dataset at a fine-scale spatial (1km) and temporal (daily) resolution for North America, Hawaii, and Puerto Rico. Model inputs are a DEM, derived horizon files, a land water mask, and daily observations from ground-based meteorological stations