Tuesday, 29 August 2023
Boundary Waters (Hyatt Regency Minneapolis)
Handout (14.7 MB)
Since the launch of the Tropical Rainfall Measuring Mission (TRMM) satellite in 1997, the follow-up satellites part of the Global Precipitation Measurement Mission (GPM) have generated till present a 26-year long precipitation data records. Over this period, the GPM constellation has been composed of two orbiting spaceborne radars and a fleet of 35 passive microwave (PMW) sensors. The GPM data archive is composed of various product levels, ranging from raw and calibrated measurements (L1) to intermediate geophysical retrieval products (L2), and multi-satellites merged datasets (L3).
The GPM-API software presented in this contribution is a python interface designed to simplify the download, processing, manipulation, and visualization of such a heterogeneous data archive which sizes several petabytes.
The GPM-API offers a user-friendly interface that enables researchers and developers to (1) download the GPM products of interest (i.e. over desired region and/or time period) from the NASA servers, (2) search for the GPM files on the local disk, (3) open the products with a single line of code in an analysis-ready-data format and making it Climate and Forecast (CF) conventions compliant, (4) facilitate on-disk (lazy) and distributed processing of the archive, (5) display the product variable(s) of interest according to their intrinsic specifications, (6) identify customly-defined precipitation events and precipitating systems across the dataset temporal record, as well as (7) extract spatial (and temporal) data patches for the design of new physical retrievals and/or the training of new machine learning algorithms.
GPM-API aims to increase scientist productivity by simplifying coding tasks associated with data download, reading, manipulation and visualization design. This will accelerate model prototyping and development, will promote the improvement of precipitation retrieval algorithms and will foster research reproducibility. These simplifications will also widen the GPM users’ community, enabling for example students and people with limited programming skills or remote sensing/meteorology background to engage with global precipitation data, making the GPM archive easily exploitable also for educational purposes.
The GPM-API software is available at https://github.com/ghiggi/gpm_api
The GPM-API software presented in this contribution is a python interface designed to simplify the download, processing, manipulation, and visualization of such a heterogeneous data archive which sizes several petabytes.
The GPM-API offers a user-friendly interface that enables researchers and developers to (1) download the GPM products of interest (i.e. over desired region and/or time period) from the NASA servers, (2) search for the GPM files on the local disk, (3) open the products with a single line of code in an analysis-ready-data format and making it Climate and Forecast (CF) conventions compliant, (4) facilitate on-disk (lazy) and distributed processing of the archive, (5) display the product variable(s) of interest according to their intrinsic specifications, (6) identify customly-defined precipitation events and precipitating systems across the dataset temporal record, as well as (7) extract spatial (and temporal) data patches for the design of new physical retrievals and/or the training of new machine learning algorithms.
GPM-API aims to increase scientist productivity by simplifying coding tasks associated with data download, reading, manipulation and visualization design. This will accelerate model prototyping and development, will promote the improvement of precipitation retrieval algorithms and will foster research reproducibility. These simplifications will also widen the GPM users’ community, enabling for example students and people with limited programming skills or remote sensing/meteorology background to engage with global precipitation data, making the GPM archive easily exploitable also for educational purposes.
The GPM-API software is available at https://github.com/ghiggi/gpm_api
- Indicates paper has been withdrawn from meeting
- Indicates an Award Winner