BrazilClim : The overcoming of limitations of pre‐existing bioclimate data

Journal Article · 07 September 2021 · International Journal of Climatology

DOI: https://doi.org/10.1002/joc.7325 · OSTI ID:1819209

Ramoni‐Perazzi, Paolo  ^[1]; Passamani, Marcelo  ^[2]; Thielen, Dirk  ^[3]; Padovani, Carlos  ^[4]; Arizapana‐Almonacid, Marco Aurelio  ^[5]

---

1. Departamento de Ecologia e Conservação Instituto de Ciências Naturais, Universidade Federal de Lavras Lavras (MG) Brazil, Modeling and Simulation Centre (CeSiMo), Faculty of Engineering University of the Andes Merida Venezuela
2. Departamento de Ecologia e Conservação Instituto de Ciências Naturais, Universidade Federal de Lavras Lavras (MG) Brazil
3. Laboratory of Landscape Ecology and Climate Venezuelan Institute for Scientific Research (IVIC) Caracas Venezuela
4. Geoprocesing Lab. Brazilian Agricultural Research Corporation (EMBRAPA—Pantanal) Corumba Brazil
5. Research Group on Remote Sensing and Mountain Ecology, School of Engineering and Environmental Management National Autonomous University of Huanta Ayacucho Peru

Abstract Species distribution modelling has become instrumental in assessing the influence of environmental conditions on the occurrence or abundance of taxa. The set of environmental layers used for this purpose is a crucial aspect, for which different climate‐based (bioclimatic) datasets have been recently developed. These bioclimatic variables result from combinations of precipitation and temperatures surfaces. Here, we explored both the performance and possibility of improving some of the currently available bioclimatic databases, through an evaluation of the precipitation and temperatures surfaces used to generate them. For this purpose, we used a combination of statistic and graphic approaches. We focused on Brazil, not only due to its natural megadiversity, but also due to its continental size and orographic heterogeneity: an excellent ground for refining methods replicable elsewhere. We found a better match between the climatic data measured on‐field and Tropical Rainfall Measuring Mission (TRMM 3B43 v7) in the case of precipitation, and the surfaces provided by the National Oceanic and Atmospheric Administration (NOAA) in the case of temperatures, sources uncommonly used for species niche modelling. We gauge‐calibrated the best performing surfaces using machine‐learning algorithms and generated corrected surfaces that allowed us to create BrazilClim: a database of bioclimatic variables, based on improved primary surfaces, which will result in more assertive predicted distributions and more actual pictures of the species' ecological requirements for megadiverse Brazil, an approach replicable elsewhere. All primary and bioclimatic surfaces generated for this study may be freely downloaded.

• Sponsoring Organization: USDOE
• OSTI ID: 1819209
• Journal Information: International Journal of Climatology, Journal Name: International Journal of Climatology Journal Issue: 3 Vol. 42; ISSN 0899-8418
• Publisher: Wiley Blackwell (John Wiley & Sons)Copyright Statement
• Country of Publication: United Kingdom
• Language: English