Abstract
To make timely decisions for weather -and climate-related disasters and
vulnerabilities, decision makers need current information that can be
readily shared and communicated to stakeholders. To date, geospatial
data is distributed using monolithic storage architectures and formats
best suited for traditional research applications. Thus, everyday
decision-makers face significant “barriers to entry” when trying to
access, explore, and modify vast historical archives and real time data
feeds. To address this need, we are developing a server architecture for
rapidly creating and publishing data products. Privileged users can
rapidly create or change products by operating on another product or
combining multiple disparate data sources together. Users can then
consume these new products using OGC-compliant WMS/WCS clients such as
ArcGIS, QGIS, or Leaflet. This enables decision makers to effectively
communicate with stakeholders using customized maps. Moreover, this
capability enables products to be rapidly updated in cases where timely
information is important. Our server architecture is containerized,
making it easy to deploy on various architectures including serverless
cloud resources. It is implemented in Python, leverages the
plug-and-play data wrangling capabilities of the PODPAC library, and
uses a custom library for serving OGC-compliant data. The result is an
easy-to-use architecture for rapidly publishing custom geospatial
products that exploit vast earth science data resources. We will
demonstrate our server capabilities by showing how privileged users can
build a set of products that are computed on-demand starting from a
fresh server. Using Jupyter Lab notebooks, we will create products that
modify single data sources as well as products that combine multiple
disparate sources. We will then show how users can consume these
products using OGC-compliant clients. Next, we will detail our
cloud-based, serverless deployment of this technology using Amazon Web
Services. Finally, we will discuss the advantages of our approach along
with any caveats. Enabling everyday decision makers to rapidly create
and share geospatial data will revolutionize their productivity and
effectiveness for assessing and remediating weather- and climate-related
vulnerabilities and disasters.