%0 Journal Article
%4 sid.inpe.br/mtc-m21c/2019/12.12.11.46
%2 sid.inpe.br/mtc-m21c/2019/12.12.11.46.24
%@doi 10.14393/rbcv70n5-45036
%@issn 0560-4613
%@issn 1808-0936
%T Reproducible geospatial data science: exploratory data analysis using collaborative analysis environments
%D 2018
%9 conference paper
%A Sánchez, Alber,
%A Vinhas, Lubia,
%A Queiroz, Gilberto Ribeiro de,
%A Simões, Rolf Ezequiel de Oliveira,
%A Gomes, Vitor,
%A Assis, Luiz Fernando Ferreira Gomes de,
%A Llapa, Eduardo,
%A Câmara, Gilberto,
%@affiliation Instituto Nacional de Pesquisas Espaciais (INPE)
%@affiliation Instituto Nacional de Pesquisas Espaciais (INPE)
%@affiliation Instituto Nacional de Pesquisas Espaciais (INPE)
%@affiliation Instituto Nacional de Pesquisas Espaciais (INPE)
%@affiliation Instituto de Estudos Avançados (IEAv)
%@affiliation Instituto Nacional de Pesquisas Espaciais (INPE)
%@affiliation Instituto Nacional de Pesquisas Espaciais (INPE)
%@affiliation Instituto Nacional de Pesquisas Espaciais (INPE)
%@electronicmailaddress albhasan@gmail.com
%@electronicmailaddress lubia.vinhas@inpe.br
%@electronicmailaddress gilberto.queiroz@inpe.br
%@electronicmailaddress rolf.simoes@inpe.br
%@electronicmailaddress vitor@ieav.cta.br
%@electronicmailaddress luizffga@dpi.inpe.br
%@electronicmailaddress edullapa@dpi.inpe.br
%@electronicmailaddress gilberto.camara@inpe.br
%B Revista Brasileira de Cartografia
%V 70
%N 5
%P 1844-1859
%K Reproducible science, data analysis, time series.
%X The answers to planetary problems could be hidden in gigabytes of satellite imagery from the last 40 years. Unfortunately, scientists lack the means for processing such amount of data as they are used to work over small quantities of satellite images. To amend this issue, we propose the use of web services from Big Earth data platforms along collaborative analysis environments. Both Web services and collaborative analysis environments fit the hypothesis-test workflow followed by researchers while writing analysis routines. Besides, the early use of Big Earth data structures eases the subsequent process of scaling analysis up to larger extensions. To test our proposal, we use our own Big Earth observation data platform, on which decades of satellite images are arranged into data cubes. By using our Web services platform, we integrate those data cubes into our collaborative analysis environment (a Jupyter notebook). Since our analysis routines consume the same data structure of the whole data sets, it is easier to scale up the analysis.
%@language en
%3 Sánchez et al. - 2018 - Reproducible geospatial data science Exploratory data analysis using collaborative analysis environments.pdf
%O “XIX Brazilian Syposium on GeoInformatics”