Weather and Climate Extremes 38 (2022) 100511

Contents lists available at ScienceDirect

Weather and Climate Extremes

journal homepage: www.elsevier.com/locate/wace

Evaluation of extreme precipitation climate indices and their projected

changes for Brazil: From CMIP3 to CMIP6
Felipe Jeferson de Medeiros a, *, Cristiano Prestrelo de Oliveira a, b, Alvaro Avila-Diaz c
a
Graduate Program in Climate Sciences, Federal University of Rio Grande Do Norte (UFRN), Natal, Rio Grande do Norte, Brazil
b
Department of Atmospheric and Climate Sciences, Federal University of Rio Grande Do Norte (UFRN), Natal, Rio Grande do Norte, Brazil
c
Faculty of Environmental and Sustainability Sciences, Universidad de Ciencias Aplicadas y Ambientales (UDCA), Bogota, Colombia

A R T I C L E I N F O A B S T R A C T

Keywords: Extreme events usually cause numerous economic and social losses, especially in vulnerable countries, such as
Climate change Brazil. Understanding whether the evolution of Earth System Models (ESMs) from Coupled Model Intercom­
Earth system models parison Project (CMIP) improves the representation of extreme events and investigating their future change is
ETCCDI . IPCC
fundamental because device policies of adaptation and mitigation to climate change generally consider the re­
Brazilian regions
sults of the most recent generation of ESMs. This study analyzes the performance of a subset of 40 ESMs from
CMIP3, CMIP5, and CMIP6 in simulating eight extreme precipitation climate indices over Brazil during
1981–2000 and also estimates their projected changes for the middle (2046–2065) and far future (2081–2 100)
under the worst-case scenario for each CMIP generation. Results reveal that CDD are the most challenging
precipitation index to be simulated, while the best ones were PRCPTOT and R20mm. The model performance
shows that CMIP3 has the best skill for Northeast Brazil, CMIP5 for Center-West, and CMIP6 for North, Southeast
and South regions. Thus, at least for Brazil, the evolution of the ESMs from CMIP did not reflect a substantial
improvement in the representation of precipitation climate extremes over all Brazilian regions. In addition, all
the models across CMIP generations have difficulty in simulating the observed trends. This indicate that im­
provements are still needed in CMIP models. Despite the relative low performance in the historical climate, the
climate projections indicate a consensus signal among most of precipitation climate extremes and CMIP gen­
erations, which increase its reliability. Overall, the extreme precipitation events are projected to be more severe,
frequent, and long-lasting in all Brazilian regions, with the more pronounce changes expected in heavy rainfall
and severe droughts in the central northern portion of Brazil and in the southern sector.

1. Introduction
Anthropogenic influence on climate change has been declared un­
doubted (IPCC et al., 2021). The latest Intergovernmental Panel on
Climate Change (IPCC) report (AR6) revealed that the occurrence of
climate extremes is linked to the observed increase in the global mean
surface air temperature (IPCC et al., 2021). The global warming in 2017
has already reached 1 ◦ C above the pre-industrial levels (IPCC et al.,
2018), mainly linked to the amount of fossil fuels burned and land
alteration by agriculture and deforestation (Ledley et al., 1999; Rajão
et al., 2020). This increase in the concentration of greenhouse gases in
the atmosphere increase downwelling infrared radiation, which en­
hances the atmospheric moisture content and intensifies the hydrolog­
ical cycle, resulting in favorable conditions for the occurrence of heavy
and extreme rainfall events (Trenberth 1999; Pfahl et al., 2017; Ayugi
et al., 2022).
Extreme rainfall events are unusual severe weather or climate con­
ditions that can potentially cause devastating impacts on humans,
infrastructure, or the environment (McPhillips et al., 2018). They are
typically defined when a rainfall value is above or below a threshold
obtained by applying statistical methods and can be short-lived, e.g.,
floods, or persist over a longer period as the droughts. According to the
Atlas of mortality and economic losses from weather, climate and water
extremes (WMO, 2019) between 1970 and 2019 there were more than
11 000 disasters were attributed to weather, climate and water-related
hazards, which accounted for just over 2 million deaths and US$ 3.64
billion losses worldwide. In Brazil, extreme weather and climate events
have been a recurrent phenomenon with considerable economic and

* Corresponding author. Federal University of Rio Grande do Norte, Campus Universitário Lagoa Nova, 59078-970 Natal, Brazil.
E-mail address: felipetkd_@hotmail.com (F. Jeferson de Medeiros).

https://doi.org/10.1016/j.wace.2022.100511
Received 21 July 2022; Received in revised form 1 September 2022; Accepted 22 September 2022
Available online 25 September 2022
2212-0947/© 2022 The Authors. Published by Elsevier B.V. This is an open access article under the CC BY license (http://creativecommons.org/licenses/by/4.0/).
F. Jeferson de Medeiros et al.
social losses. During the last decade (2010–2019) remarkable natural
disasters related to extreme rainfall episodes have affected Brazil. For
example, the severe droughts in Amazonia in 2010, 2015–2016 (Lewis
et al., 2011; Marengo et al., 2011; Jiménez-Muñoz et al., 2016); in the
Northeast during 2012–2017 (Brito et al., 2017; Medeiros et al., 2020a);
in Southeast in 2014 (Coelho et al., 2016; Nobre et al., 2016), and in the
Brazilian Pantanal in 2019 (Libonati et al., 2020; Marengo et al., 2021).
Likewise, notable extreme floods were reported such as the floods in
2012 and 2014 in Amazonia (Espinoza et al., 2013, 2022), and the
landslides associated with the flood event in Rio de Janeiro state in 2011
which killed almost 1000 people (Busch and Amorim, 2012). Solely the
Northeast Brazil drought event (2012–2017) affected almost 33.4
million people, and caused an estimated loss of US $30.0 billion
(Marengo et al., 2017a).
In the begging of the 2020s-decade extreme rainfall events continue
to affect the Brazilian territory. For example, urban floods were recorded
in the states of Bahia and Minas Gerais in 2021 (Cortez et al., 2022), and
in 2022 extreme rainfall hit the city of Petropolis in Rio de Janeiro and
the metropolitan area of Recife, causing extensive damage, several
homeless and approximately 400 deaths (Zachariah et al., 2022). In
addition, a new historical flood was reported in the Amazon Basin in
June 2021 (Espinoza et al., 2022). Although all these negative
socio-economic impacts and some observational studies have shown
evidence of upward trends in extreme precipitation in many areas of
Brazil (Teixeira and Satyamurty, 2011; Ávila et al., 2016; Zilli et al.,
2017; Costa et al., 2020; Marengo et al., 2020a; Regoto et al., 2021), few
studies have addressed how extreme rainfall events may behave in the
future climate, especially focusing on the whole Brazilian territory.
Extreme rainfall events have been usually investigated using the
Expert Team on Climate Change Detection and Indices (ETCCDI) (Zhang
et al., 2011). The ETCCDI is a set of internationally-accepted indices
based on daily air temperature and precipitation measures that eluci­
dates changes in the frequency, duration and magnitude of extreme
climate events (Donat et al., 2013; Sillmann et al., 2013a, 2013b).
Several researchers worldwide have applied the ETCCDI climate indices
to analyze possible changes in extreme precipitation events over the past
(Brown et al., 2010; Sillmann et al., 2013a; Donat et al., 2016; lacovone
et al., 2020; Cerón et al., 2021; Regoto et al., 2021; Faye and Akinsanola,
2022) and for future climate using climate models projections (Sillmann
et al., 2013b; Thibeault and Seth, 2014; Xu et al., 2019; Avila-Diaz et al.,
2020a; Gouveia et al., 2022).
In this perspective, one of the possibilities to approach and under­
stand how extreme rainfall events may behave in the future climate are
using the climate simulations and projections of the Earth System
Models (ESMs) from the Coupled Model Intercomparison Project
(CMIP), which are the main sources used in the assessment reports of the
IPCC. Three CMIP model generations are currently available: CMIP3
(Meehl et al., 2007), CMIP5 (Taylor et al., 2012) and CMIP6 (Eyring
et al., 2016). These models have been constantly updated to systemat­
ically advance our understanding of the complete Earth system across
CMIP generations. For example, compared to CMIP3, the CMIP5 ESMs
include interactive ocean and land carbon cycles, more comprehensive
modeling of the indirect effect of aerosols, and the use of time-evolving
volcanic and solar forcing in most models (Taylor et al., 2012; Knutti and
Sedlácek 2013). For CMIP6, improvements are expected mainly for the
methods of initialization and ensemble generation (Eyring et al., 2016).
Moreover, the CMIP6 models include higher spatial resolution, a wide
range of scenarios, higher climate sensitivity, and additional physical
complexity (e.g., biogeochemistry) relative to previous CMIP versions
(Bock et al., 2020).
Even though this continuous improvement of ESMs and some studies
have shown that the evolution across CMIP generations has reduced
model uncertainties for some areas (Blázquez and Nuñez, 2013; Zhu
et al., 2020; Srivastava et al., 2020; Gusain et al., 2020; Ukkola et al.,
2020; Luo et al., 2021), other studies have shown that the evolution of
the CMIP model generations did not improve the representation of
2
Weather and Climate Extremes 38 (2022) 100511
climate features, especially the simulation of extreme precipitation
climate indices (Agel and Barlow, 2020a; Akinsanola et al., 2020;
Wehner et al., 2020; Seneviratne and Hauser, 2020). It is widely known
that ESMs may not adequately reproduce extreme precipitation events
and sub-grid processes due to their usually lower resolution. However,
Bador et al. (2020) show that increasing spatial resolution alone is
insufficient to obtain a systematic improvement in the simulation of
precipitation extremes. Therefore, studies investigating whether the
evolution of CMIP models improves the representation of the current
climate are vital because the newest IPCC reports generally consider the
results of the most recent generation of ESMs.
In this way, using the simulations and climate projections of the
ESMs from the CMIP and the extreme climate indices recommended by
the ETCCDI is an excellent tool to investigate extreme precipitation
climate events and their projected changes, especially for Brazil that
needs to advance in studies on climate modeling with the objective of
improve knowledge about climate change and its impacts over the
whole country. This lack of studies is verified even knowing that Brazil is
highly vulnerable to climate change, mainly in its tropical area (Torres
and Marengo, 2013; Darela-Filho et al., 2016; Debortoli et al., 2017). For
example, Torres and Marengo (2013) showed that the increase in tem­
perature might exceed 4 ◦ C by the end of the 21st century over the
Amazon basin and Northeast Brazil. Avila-Diaz et al. (2020a) showed
that the number of Consecutive Wet Days (CWD) could decrease by 10%
in most of the Amazon watershed basin by mid-century (2046–2065).
Medeiros and Oliveira (2022) showed that the number of dry days and
heavy rainfall days are projected to increase up to 15% and can exceed
100%, respectively, over Northeast Brazil in the far future (2076–2 100).
Therefore, given the lack of studies assessing whether the evolution
of the ESMs from CMIP3 to CMIP6 improves the representation of
extreme climate events over Brazil and considers the remaining need for
more in-depth information on future climate in South America, this
study has two main objectives: (1) evaluate the performance of a subset
of ESMs from CMIP3, CMIP5, and CMIP6 in representing extreme pre­
cipitation climate indices during the historical period in Brazil
(1981–2000); and (2) estimate how they may change in the middle
(2046–2065) and far future (2081–2 100) under the worst-case scenario
for each CMIP generation.
2. Data and methods
2.1. Study area
The study area is Brazil (Fig. 1a and b), which covers an area of
approximately 8.516 000 km2 with a population of over 214 million
inhabitants (IBGE 2021). The country has important physical charac­
teristics and natural resources, such as the largest forest on the planet
(Amazon rainforest), the most densely populated dry land region in the
world (semiarid region located in Northeast Brazil), one of the largest
wetlands on earth (the Pantanal region), and finally the fifth largest
watershed in the world (La Plata basin) (Alvalá et al., 2017; Zanin and
Satyamurty, 2020; Libonati et al., 2020). Brazil is also one of the world’s
largest exporters of agriculture and livestock, economic activities
strongly influenced by weather and climate extremes. Most of the energy
mix in Brazil is from hydropower, contributing more than 60% of its
total installed electric generation capacity (Luiz-Silva and Garcia, 2022).
Due to the vast territorial extension, the precipitation regimen in
Brazil is modulated by many atmospheric systems (Luiz-Silva et al.,
2021). These atmospheric conditions provide rainy and dry seasons at
different times throughout the year in Brazil’s five macro regions
(Fig. 1b): North (NT), Northeast (NE), Center-West (CW), Southeast
(SE), and South (ST). However, in general most of Brazil is characterized
by a monsoon climate, with a predominance of the wet (dry) season
during the austral summer and autumn (winter and spring), mainly
associated with the meteorological systems of the South Atlantic
Convergence Zone (SACZ) (Carvalho et al., 2004; Pezzi et al., 2022) and
F. Jeferson de Medeiros et al. Weather and Climate Extremes 38 (2022) 100511

Fig. 1. Geographical location of Brazil (a) with its five macro regions highlighted (b). The acronym (1), (2), (3), (4) and (5) refers to the Brazilian regions: North,
Northeast, Center-West, Southeast and South, respectively.

Intertropical Convergence Zone (ITCZ) (Uvo et al., 1998; Medeiros et al., (CDD) and consecutive wet days (CWD). The calculated indices allowed
2020b). Details about the precipitation regimen for each Brazil macro the assessment of intensity (PRCPTOT, R95p, RX1day, RX5day and
region (Grimm et al., 1998; Borges et al., 2014; Santos et al., 2015; 2021; SDII), frequency (R20mm), and duration (CDD and CWD) of extreme
Oliveira et al., 2017) and its main atmospheric systems (Kousky, 1979; precipitation climate events and were chosen because they represent key
Cohen et al., 1995; Gomes et al., 2019; Reboita et al., 2019; Montini extreme events that mostly affect the study area (Debortoli et al., 2017).
et al., 2019; Schumacher and Rasmussen, 2020; Reis et al., 2021; Jesus Moreover, these indices have been widely used to study observed and
et al., 2021) can be found at the references mentioned above. modeled climate variability across the globe (Sillmann et al., 2013a,
2013b; Avila-Diaz et al., 2020a, 2020b; Faye and Akinsanola, 2022;
Gouveia et al., 2022; Wilson et al., 2022).
2.2. Extreme precipitation climate indices To improve the inter-annual extreme precipitation variability across
the Brazilian regions and considering that the extreme climate indices
Eight extreme precipitation climate indices defined by ETCCDI were are more meaningful on annual scale (Aerenson et al., 2018; Avila-Diaz
selected for this study (Table 1): total wet-day precipitation (PRCPTOT), et al., 2020b), the selected extreme precipitation climate indices have
very wet day (R95p), maximum 1-day precipitation (RX1day), been calculated annually.
maximum 5-day precipitation (RX5day), simple daily intensity (SDII),
number of heavy precipitation days (R20mm), consecutive dry days
2.3. Observation database
Table 1
Extreme precipitation climate indices used in this study as recommended by the The observational dataset used in this study was provided by Xavier
ETCDDI. et al. (2016). The database contains daily precipitation information and
was developed using the inverse distance weighting interpolation
Index Index name Index definitions Units
method applied to 3625 rain gauges and 735 weather stations covering
PRCPTOT Annual wet-day Annual total precipitation in wet mm the entire Brazilian territory, providing data with a spatial resolution of
precipitation days a
R95p Very wet days Annual total precipitation from mm
0.25◦ × 0.25◦ during the 1980–2016 period (https://utexas.app.box.co
days >95th percentile m/v/Xavier-etal-IJOC-DATA).
RX1day Max 1-day Annual maximum 1-day mm In addition to Xavier et al. (2016) there are several other data sources
precipitation precipitation in the literature that could be used to evaluate the performance of the
RX5day Max 5-day Annual maximum consecutive 5- mm
ESMs across CMIP generations, such as Global Meteorological Forcing
precipitation day precipitation
SDII Simple daily The ratio of annual total mm/ Dataset (GMFD) (Sheffield et al., 2006), Multi-Source Weight­
precipitation index precipitation to the number of day ed-Ensemble Precipitation (MSWEP) (Beck et al., 2019), Climate Haz­
wet days a ards Group InfraRed Precipitation with Station data (CHIRPS) (Funk
R20mm Number of very heavy Annual count of days when daily days et al., 2015), Global Precipitation Climatology Project One-Degree Daily
precipitation days precipitation ≥20 mm
CDD Consecutive dry days Maximum number of consecutive days
Dataset (GPCP) (Huffman et al., 2001). However, some studies have
dry days a shown that Xavier database provided a better representation of the
CWD Consecutive wet days Maximum number of consecutive days Brazilian rainfall (Avila-Diaz et al., 2020a, 2020b; Lucas et al., 2021),
wet days a mainly due to quality control, number of station networks in their
More details on definitions of the core indices given at https://etccdi.pacificclim development and for not taking into account reanalysis product and
ate.org/list_27_indices.shtml satellite rainfall estimation. Furthermore, this dataset has already been
a
The ETCDDI defined a wet (dry) day when precipitation ≥1 mm (<1 mm). used in several applications in Brazil, including the analysis of severe

3
F. Jeferson de Medeiros et al. Weather and Climate Extremes 38 (2022) 100511

drought events (Medeiros et al., 2020a; Medeiros and Oliveira, 2021), Table 2
extreme precipitation climate indices (Silva et al., 2019; Lucas et al., Details of the 40 ESMs used in the present study. The character “*” specify un­
2021), performance of regional climate model (Medeiros et al., 2020b) availability for future periods under the respective scenario.
and potential changes in surface solar radiation (Zuluaga et al., 2022). CMIP Model Institution, Resolution Runs/ Experiment
Country (Lat◦ X Ensemble protocol
Lon◦ )
2.4. Simulations and climate projections (CMIP)
CMIP3 01. CNRM- CNRM, 2.8◦ × 2.8◦ r1 Historical +
In order to calculate the extreme climate indices, a survey was CM3 France SRES A2
02. FGOALS, 2.8◦ × 2.8◦ r1 Historical +
initially carried out to identify which “families” of models are present
FGOALS- China SRES A2*
from CMIP3 to CMIP6 and has data available on a daily scale. After 1-0-G
carrying out this survey, it was found that there are 7 families of models 03. GFDL- GFDL, USA 2.5◦ × 2.0◦ r1 Historical +
that meet the criteria mentioned above: Center National de Recherches CM2-0 SRES A2
04. GFDL- GFDL, USA 2.5◦ × 2.5◦ r2 Historical +
Meteorologiques (CNRM), Institute of Atmospheric Physics (FGOALS),
CM2-1 SRES A2
Geophysical Fluid Dynamics Laboratory (GFDL), Institute Pierre-Simon 05. IPSL- IPSL, 3.8◦ × 2.5◦ r1 Historical +
Laplace (IPSL), Atmosphere and Ocean Research Institute (MIROC), Max CM4 France SRES A2
Planck Institute for Meteorology (MPI) e Meteorological Research 06. MIROC, 1.1 × 1.1
◦ ◦
r1 Historical +
Institute (MRI). After identifying the model families present in all CMIP MIROC3- Japan SRES A2*
2-HIRES
versions, all available model sets were selected. With this step, 40
07. MIROC, 2.8◦ × 2.8◦ r1 Historical +
models were identified (Table 2), 9 belonging to CMIP3 (Meehl et al., MIROC3- Japan SRES A2
2007), 16 to CMIP5 (Taylor et al., 2012), and 15 to CMIP6 (O’Neil et al., 2-MEDRES
2016). 08. MPI- MPI, 1.9◦ × 1.9◦ r1 Historical +
ECHAM5 Germany SRES A2
After selecting the models, the next step was selecting the years of
09. MRI- MRI, Japan 2.8◦ × 2.8◦ r1 Historical +
study. It was found that the historical period available for each version CGMC2-3- SRES A2
of the ESMs that are part of CMIP3 (1961–2000), CMIP5 (1850–2005), 2A
and CMIP6 (1850–2014) are different. Thus, as one of the study’s ob­ CMIP5 10. CNRM- CNRM, 1.4◦ × 1.4◦ r1i1p1 Historical +
jectives is precisely to assess whether the evolution of the CMIP models CM5 France RCP8.5
11. FGOALS, 2.8 × 2.8
◦ ◦
r1i1p1 Historical +
resulted in improvements in the representation of extreme rainfall
FGOALS- China RCP8.5
events, it was decided to select a period compatible with the three G2
versions. Following this criterion, one of the possible periods was 12. FGOALS, 2.8◦ × 2.8◦ r1i1p1 Historical +
1961–2000. However, the observational dataset is available for the FGOALS- China RCP8.5
1980–2016 period. Thus, the period 1981–2000 was selected to be used S2
13. GFDL- GFDL, USA 2.5◦ × 2.0◦ r1i1p1 Historical +
to analyze the historical climate. CM3 RCP8.5
In order to understand how extreme rainfall events may behave in 14. IPSL- IPSL, 3.8 × 1.9
◦ ◦
r1i1p1 Historical +
the future, two-time slices were considered: 2046–2065 (mid-century CM5A-LR France RCP8.5
future) and 2081–2 100 (far future). The climate projections throughout 15. IPSL- IPSL, 2.5◦ × 1.3◦ r1i1p1 Historical +
CM5A-MR France RCP8.5
the 21st century with the sets of ESMs that are part of CMIP3, CMIP5,
16. IPSL- IPSL, 3.8◦ × 1.9◦ r1i1p1 Historical +
and CMIP6 were produced considering different prescribed forcing CM5B-LR France RCP8.5
agents and distinct hypotheses about population growth, economic 17. MIROC, 2.8 × 2.8
◦ ◦
r1i1p1 Historical +
development, energy model, globalization, among other factors. Thus, MIROC- Japan RCP8.5
the direct comparison between the CMIP generations scenarios is not a ESM
18. MIROC, 2.8◦ × 2.8◦ r1i1p1 Historical +
trivial activity, as highlighted by Rogelj et al. (2012) and Sillmann et al. MIROC- Japan RCP8.5
(2013b). However, taking into consideration the current greenhouse gas ESM-
emission trends that recently reached 421 ppm of CO2 – 50% higher than CHEM
the preindustrial levels – and harmful environmental decision-making 19. MIROC, 0.6◦ × 0.6◦ r1i1p1 Historical +
MIROC4H Japan RCP8.5*
by many countries, such as Brazil (Artaxo 2019), the most probable
20. MIROC, 1.4◦ × 1.4◦ r1i1p1 Historical +
climate scenario to happen in the future is the most pessimistic. In this MIROC5 Japan RCP8.5
perspective, the climate projections will be analyzed considering the 21. MPI- MPI, 1.9◦ × 1.9◦ r1i1p1 Historical +
SRES A2, RCP8.5, and SSP5-8.5 scenarios for CMIP3, CMIP5, and ESM-LR Germany RCP8.5
CMIP6, respectively. The SRES A2 emission scenario corresponds to an 22. MPI- MPI, 1.9◦ × 1.9◦ r1i1p1 Historical +
ESM-MR Germany RCP8.5
equivalent CO2 concentration of approximately 850 ppm in 2 100 23. MPI- MPI, 1.9 × 1.9
◦ ◦
r1i1p1 Historical +
(Nakicenovic et al., 2000). The RCP8.5 corresponds to an approximate ESM-P Germany RCP8.5*
radiative forcing by the end of the century – relative to preindustrial 24. MRI- MRI, Japan 1.1◦ × 1.1◦ r1i1p1 Historical +
conditions – of 8.5 Wm− 2, and equivalent CO2 concentrations roughly CGMC3 RCP8.5
25. MRI- MRI, Japan 1.1◦ × 1 × r1i1p1 Historical +
to 1370 ppm (Moss et al., 2010). Finally, the SSP5-8.5 is characterized
ESM1 1◦ RCP8.5
by potential future fossil-fueled growth (O’Neil et al., 2016). CMIP6 26. CNRM- CNRM, 1.4◦ × 1.4◦ r1i1p1f2 Historical +
The ESMs spatial resolutions vary from roughly 1–4◦ of latitude/ CM6 France SSP5-8.5
longitude (Table 1). However, for intercomparison purposes all datasets 27. CNRM- CNRM, 0.5◦ × 0.5◦ r1i1p1f2 Historical +
(observation and models) were interpolated to a common 1.0◦ × 1.0◦ CM6-1-HR France SSP5-8.5
28. CNRM- CNRM, 1.4◦ × 1.4◦ r1i1p1f2 Historical +
grid using a first-order conservative remapping technique (Jones, 1999), ESM2-1 France SSP5-8.5
following analogous studies (e.g., Torres and Marengo, 2013; Lovino 29. FGOALS, 1.3◦ × 1.0◦ r1i1p1f1 Historical +
et al., 2018, 2021; Almazroui et al., 2021). Regarding the choice of FGOALS- China SSP5-8.5*
members, some studies have already shown that simulations with mul­ F3-L
30. FGOALS, 2.0◦ × 2.0◦ r1i1p1f1 Historical +
tiple members reduce the internal variability of the climate system and
FGOALS- China SSP5-8.5
model uncertainties (Lehner et al., 2020; Deser et al., 2020; Díaz et al., G3
2021; Lovino et al., 2021). However, taking into account mainly that (continued on next page)
most of the climate simulations and projections of the CMIP3 models

4
F. Jeferson de Medeiros et al. Weather and Climate Extremes 38 (2022) 100511

Table 2 (continued ) ∑n
⎧ |mi − Oi | ∑n ∑n
CMIP Model Institution, Resolution Runs/ Experiment ⎪
⎪ 1 − ∑ i=1
, when |mi − Oi | ≤ 2x |Oi − Oi |
⎪
⎪ n
Country (Lat◦ X Ensemble protocol ⎨ 2x i=1 |Oi − O| i=1 i=1
Lon◦ ) dr = (03)
⎪ ∑n
⎪ ∑n ∑n
31. GFDL- GFDL, USA 1.3◦ × 1.0◦ r1i1p1f1 Historical + ⎪ 2x
⎪
⎩ ∑n i=1
|Oi − O|
− 1, when |mi − Oi | > 2x |Oi − Oi |
ESM4 SSP5-8.5 |mi − Oi | i=1 i=1
32. IPSL- IPSL, 3.8◦ × 1.9◦ r1i1p1f1 Historical + i=1

CM5A2- France SSP5-8.5* ∑n

INCA i=1 (mi − mi )(Oi − Oi )
CORR = √̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅
̅ √∑ (04)
33. IPSL- IPSL, 2.5◦ × 1.3◦ r1i1p1f1 Historical + ∑n ̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅
n
̅
CM6A-LR France SSP5-8.5 i=1 (Oi − Oi ) i=1 (mi − mi )
34. IPSL- IPSL, 2.5◦ × 1.3◦ r1i1p1f1 Historical +
CM6A-LR- France SSP5-8.5* where Oi is the observed value, Oi is the mean of observed data, mi is the
INCA
simulated value, mi is the mean of simulated, e n is the total observation
35. MIROC, 2.8 × 2.8
◦ ◦
r1i1p1f2 Historical +
MIROC- Japan SSP5-8.5 number.
ES2L PBIAS indicates whether a given dataset overestimates or un­
36. MIROC, 1.4◦ × 1.4◦ r1i1p1f1 Historical + derestimates the observational information. RSR is calculated as the
MIROC6 Japan SSP5-8.5
ratio of the RMSE and standard deviation of observed data (Moriasi
37. MPI- MPI, 1.9◦ × 1.9◦ r1i1p1f1 Historical +
ESM-1-2- Germany SSP5-8.5*
et al., 2007). The closer PBIAS and RSR are to 0, the better is the model
HAM performance. The refined index of agreement (dr) varies between − 1
38. MPI- MPI, 0.9◦ × 0.9◦ r1i1p1f1 Historical + and 1. A dr of 1 indicates a perfect agreement and dr = − 1 indicates
ESM-1-2- Germany SSP5-8.5 either a lack of agreement between observed and simulated values or a
HR
lack of variability in the observed data (Willmott et al., 2015). Finally,
39. MPI- MPI, 1.9◦ × 1.9◦ r1i1p1f1 Historical +
ESM-1-2- Germany SSP5-8.5 the value of CORR 1 (− 1) indicates a stronger positive (negative) rela­
LR tionship between the two variables; meanwhile, the 0 value indicates
40. MRI- MRI, Japan 1.1◦ × 1.1◦ r1i1p1f1 Historical + there is a lack of any linear relationship between observed (Oi) and
ESM2-0 SSP5-8.5
simulated (mi) data.
In order to select the best models that better represent the extreme
have only one run available, and following the coding established by the precipitation climate indices over Brazil and then assesses the climate
WCRP working group on coupled modeling and also some studies in the projections with a more robust signal, the comprehensive model rank
literature (e.g., Sillmann et al., 2013b; Lovino et al., 2018; Agel and (MR ) (Eq. (05)) had also been calculated (Jiang et al., 2015; Zhang et al.,
Barlow, 2020b; Wild, 2020; Alves et al., 2021) here we used one or two 2018). MR is based on the performance metrics of PBIAS, RSR, dr, and
model run experiment and single-member simulations. As this study CORR and is a measure of how consistently each model is classified
proposes a first-order assessment of extreme precipitation climate among all the ranking possibilities (indices and metrics). The maximum
indices intercomparison across the three CMIP generations for Brazil, value of MR is 1, indicating that the model is the best in all indices and
further works with multiples ensemble can be done to investigate these metrics.
possible outcomes differences. 1 ∑ n
( )
Lastly, besides analyzing the extreme climate indices by each model MR = 1 − RankiPBIAS + RankiRSR + Rankidr + RankiCORR (05)
n x m i=1
separately, we test whether using Multi-Model Ensembles (MMEs) can
improve the representation of extreme precipitation climate indices.
where n is the total number of indices, m is the number of models and the
Calculating the ensemble mean is a common technique for avoiding the
Ranki indicates the order of each CMIP model on each index in a given
large spread found in individual model results (Sillmann et al., 2013a;
performance metric.
Avila-Diaz et al., 2020a, 2020b). For that end, the mean of each index
To detect trends in extreme climate indices, we used Theil-Sen’s
among the CMIP generation (CMIP3, CMIP5 and CMIP6) models were
slope estimator (Sen, 1968). In addition, the significance of trends is
calculated and we treated those as separate results. All the analyses were
calculated at the confidence level of 95% (α = 0.05) using a Man­
done for Brazil and the five Brazilian macro regions (Fig. 1b). The
n–Kendall test (Mann, 1945; Kendall, 1975). A full discussion of the
subdivision was performed to account for its heterogeneity in climate
Mann Kendall and Sen slope methods can be found in Yue et al. (2002).
extremes.
These nonparametric tests have been broadly used for detecting trends
in extreme rainfall events (Oliveira et al., 2017; Avila-Diaz et al., 2020a,
2.5. Performance and trend calculation
2020b; Wilson et al., 2022; Medeiros and Oliveira, 2022).

Four statistical metrics were employed to evaluate the performance

2.6. Future changes
of the simulated extreme precipitation climate indices over the five
macro regions of Brazil: Percent Bias (PBIAS) (Eq. (01)), RMSE-obser­
To assess the changes between the future and the historical periods,
vations standard deviation ratio (RSR) (Eq. (02)), refined index of
the relative change in each climate extreme index (CEI) was calculated
agreement (dr) (Eq. (03)) and the Pearson correlation coefficient
using Eq. (06) adapted from Bador et al. (2018):
(CORR) (Eq. (04)).
∑n CEI future − CEI hist
(mi − Oi ) x 100 Relative change in CEI = ∗ 100 (06)
PBIAS = i=1 ∑n (01) CEI hist
i=1 Oi

√̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅ where CEIfuture e CEIhist are 20-yr averages in a given CEI over the future
RMSE (mi − Oi )2 (2046–2065 or 2081–2 100) and historical (1981–2000) periods,
RSR = = √̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅ (02)
STDEVObs respectively.
(Oi − Oi )2

5
F. Jeferson de Medeiros et al.
3. Results and discussion
3.1. Evaluation of CMIP ESMs in representing extreme precipitation
climate indices
3.1.1. Observed climatology
Fig. 2 shows the spatial distribution of the analyzed eight extreme
precipitation climate indices calculated using the database from Xavier
et al. (2016). The extreme events associated with an excess of rainfall
(PRCPTOT, RX1day, RX5day, SDII, R20mm, CWD) are more intense,
frequent and long-lasting in the North (NT), Southeast (SE) and South
(ST) regions of Brazil. Despite this common aspect, it is important to
highlight that depending on the extreme precipitation climate index
investigated, the pattern between the regions are relatively different. For
example, in the NT the PRCPTOT (Fig. 2a) and CWD (Fig. 2h) stand out
in comparison with the other regions. In this case, the PRCPTOT values
are greater than 1800 mm, and the number of CWD are higher than 100
days over most of the region. This is mainly due to the advection of
moisture from the North Atlantic Ocean and the evapotranspiration
from the Amazon rainforest, which supply atmospheric moisture and
contributes to higher annual precipitation amounts and a higher
occurrence of rainy days (Hasler and Avissar, 2007; Drumond et al.,
2014; Builes-Jaramillo et al., 2018; Funatsu et al., 2021). In the SE and
ST, the precipitation climate indices of R95p (Fig. 2b), RX1day (Fig. 2c)
and RX5day (Fig. 2d) stands out, in which are observed values of
RX1day and RX5day greater than 100 mm and 240 mm, respectively,
over almost the entire South region and in the southeastern portion of
SE. These extreme events are related to heavy and persistent rainfall,
and are possibly influenced by the actions of the frontal zone, CCM, and
SACZ, which are the meteorological systems that usually cause the
largest accumulation of precipitation in these regions (Carvalho et al.,
2002, 2004; Durkee et al., 2009; Teixeira and Satyamurty, 2011; Aguiar
and Cataldi, 2021).
In Northeast Brazil (NE), the extreme precipitation climate events
are less intense and frequent. In this region, the low simple daily pre­
cipitation (SDII) (Fig. 2e) and the high number of Consecutive Dry Days
(CDD) stand out, which according to Fig. 2g exceeds 40% of the annual
days in the central portion of NE, area with semi-arid climate. This
pattern is directly associated with low annual total rainfall and short wet
Fig. 2. Spatial distribution of the eight extreme precipitation climate indices over Brazil during 1981–2000: (a) annual wet-day precipitation (PRCPTOT), (b) very
wet days (R95p), (c) annual maximum 1-day precipitation (RX1day), (d) annual maximum consecutive 5-day precipitation (RX5day), (e) simple daily intensity
(SDII), (f) Annual count of days when daily precipitation ≥20 mm (R20mm), (g) number of consecutive dry days (CDD) and (h) number of consecutive wet days
(CWD). The precipitation dataset used is from Xavier et al. (2016).
6
Weather and Climate Extremes 38 (2022) 100511
season (Medeiros et al. 2020a, b; Medeiros and Oliveira, 2021). On the
other hand, in the Center-West (CW), the behavior varies depending on
the extreme precipitation climate indices investigated. For some indices,
such as PRCPTOT, R95P, SDII, R20mm e CWD the northern portion
shows extreme events more intense, frequent and long-lasting, while for
RX1day and RX5day the southern portion is the area in which the
extreme events are more intense.
3.1.2. Model performance for Brazilian regions
The performance of the 40 ESMs from CMIP3, CMIP5, and CMIP6
and their Multi-Model Ensemble (MMEs) for the North Brazil region is
evaluated in Fig. 3. Most models underestimate the precipitation indices
in the region (Fig. 3a), except for CDD and CWD. This underestimation
of extreme precipitation is also verified when analyzing the annual and
seasonal cycle of precipitation, as pointed out by Torres and Marengo
(2013), Jones and Carvalho (2013), Gulizia and Camilloni (2015), and
Alves et al. (2021). According to Torres and Marengo (2013) the sys­
tematic rainfall underestimation over the Amazon basin by the global
circulation models may be related to numerous factors, such as the poor
representation of cumulus convection, the biosphere-atmosphere in­
teractions in the rainforest, soil moisture, and land surface processes.
Moreover, few data observation coverage is available for the region,
which influences the magnitude and location of precipitation. The
lowest performance of the ESMs is recorded for the RX1day, RX5day and
CDD indices (Fig. 3a and b), while the better are detected for PRCPTOT,
R20mm and SDII (Fig. 3b and c), as also found by Avila-Diaz et al.
(2020b). Regarding the performance of the model families, we highlight
the good performance of CNRM in the CMIP6 version, the MPI in CMIP5,
and the MRI in CMIP5 and CMIP6 (Fig. 3a–d). On the other hand, an
inadequate performance was found by the FGOALS and IPSL models in
CMIP3 and MIROC in CMIP6.
For the Northeast Brazil region (Fig. 4), the errors in the estimation
of precipitation indices remain high in most models (Fig. 4a). However,
unlike was previously found for NT (Fig. 3a), in the NE the eight extreme
precipitation climate indices that are intensity, frequency and duration-
related depicts a predominance of overestimation (PBIAS >0), as is
usually observed when the rainfall is analyzed (Marengo et al., 2017b;
Ortega et al., 2021). Similarly for the NT region, it is noted that the
PRCPTOT, R20mm, and SDII continue to be the better precipitation
F. Jeferson de Medeiros et al. Weather and Climate Extremes 38 (2022) 100511

Fig. 3. Statistical performance for annual extreme precipitation climate indices for individual CMIP3 (blue), CMIP5 (red) and CMIP6 (black) models, and their multi-
model ensembles (MME) from 1981 to 2000 over the North Brazil region. The statistical metrics refers to (a) Percent Bias (PBIAS); (b) RMSE-observations standard
deviation ratio (RSR); (c) a refined index of model performance (dr), and (d) Correlation coefficient (CORR). (For interpretation of the references to colour in this
figure legend, the reader is referred to the Web version of this article.)

Fig. 4. As in Fig. 3, but for Northeast Brazil region.

climate indices simulated by the models (Fig. 4a–d). The indices in
which the models show the worst representation skill are also similar to
those found for NT region. In this case, in addition to RX1day, RX5day
and CDD, the CWD are also not well simulated. Regarding the perfor­
mance of the model families, the better skills are found by FGOALS in
CMIP5 and CMIP6, MIROC in CMIP3 and CMIP5, and MPI in CMIP6. On
the other hand, the CMIP5 versions of GFDL and MIROC show notice­
ably lower performance compared to their versions of CMIP3 and CMIP6
(Fig. 4c).
In the Center-West (Fig. 5), despite the model’s performance being
relatively better compared to the NT (Fig. 3) and NE (Fig. 4) regions, the
RX1day, CDD, and CWD are again the extreme precipitation climate
indices more difficult to represent. A special emphasis is given for CDD,
in which all models indicate overestimation (Fig. 5a) and lack of
agreement between simulated and observed values (Fig. 5c). The best
performances are found for RX5day, R20mm and SDII.
The analysis of the evolution of the models indicates that in some
cases, such as for the GFDL and MIROC, there were apparently no

7
F. Jeferson de Medeiros et al. Weather and Climate Extremes 38 (2022) 100511

Fig. 5. As in Fig. 3, but for Center-West Brazil region.

improvements in the simulation of precipitation climate extremes over
the Center-West, given that the performance of the models was better in
CMIP3 compared to CMIP5 and CMIP6 (Fig. 5a–d). On the other hand,
the FGOALS-G3 of CMIP6 presents better performance compared to the
other model versions. Therefore, depending on the subset of models
used, the evolution of the CMIP model does not apparently mean
improvement in the simulating of extreme precipitation events, as
observed for Brazil’s North, Northeast and Center-West regions
(Figs. 3–5). However, since other aspects such as spatial variability
could be improved, these results must be carefully analyzed. Neverthe­
less, although the cautions are needed, they are in agreement with what
was found by Kim et al. (2020) for South America, and Medeiros and
Oliveira (2022) for Northeastern Brazil, which indicates reasonable
usability in the results presented here. Such a result is verified even
though several studies (e.g., Sperber et al., 2013; Gusain et al., 2020;
Arias et al., 2021) have shown that there have been improvements in the
simulation of precipitation at regional scale across CMIP generations,
mainly due to the most appropriate representation of sub-grid scale
process, such as convection and cloud microphysics (Neale et al., 2008;
Wu et al., 2019; Wyser et al., 2020).
For the SE and ST regions (Figs. 6 and 7) the errors in the estimation
of precipitation indices are lower when compared to the NT (Fig. 3), NE
(Fig. 4), and CW regions (Fig. 5), corroborating the results found by
Marengo et al. (2021). A possible explanation for this better perfor­
mance in these regions may be related to the geographical location. As
they are located in southern latitudes, near the mid-latitudes, these

Fig. 6. As in Fig. 3, but for Southeast Brazil region.

8
F. Jeferson de Medeiros et al.
Fig. 7. As in Fig. 3, but for South Brazil region.
regions have an atmosphere predominantly baroclinic in which the
dynamical equations are better solved (Bador et al., 2020). However, for
both regions, the precipitation indices of R20mm, CDD, and CWD are the
most difficult to represent. This indicates that for all regions in Brazil the
number of Consecutive Dry Days (CDD) is the most challenging pre­
cipitation index to be simulated, a result similar to that found for other
locations worldwide, such as China (Luo et al., 2021). In this sense, it is
important to know that in almost all cases the models overestimate the
CDD index (Figs. 3–7). This indicates that the models sign drier condi­
tions than usually observed. On the other hand, the precipitation indices
of PRCPTOT and SDII are well simulated. This could generate an un­
derestimation on heavy rainfall indices; for this reason, care is necessary
for the R95p and RX5day in these regions.
In addition to the lower estimation errors for the SE and ST regions,
another interesting point is that the evolution of the CMIP models in
most cases improved the representation of precipitation indices, unlike
were observed for the other regions of Brazil. For the Southeast region
(Fig. 6), from the 7 model families used in 4 the CMIP6 models showed
superior performance compared to the CMIP3 and CMIP5 versions
(CNRM, FGOALS, IPSL, MRI). In the South region (Fig. 7), the
improvement of the models with their evolution is even more visible. In
this case, only the MPI versions of CMIP6 did not perform better when
compared to their previous versions. This result again reinforces how
CMIP models appear to be better configured to represent extreme
rainfall events in a baroclinic atmosphere. Overall, the results suggest
that there is no consensus among all regions of Brazil if the improve­
ments throughout the CMIP versions of the models have improved the
simulation of precipitation climate extremes. Thus, the statistical metric
of model rank will be presented in the next section to analyze this aspect
in more detail.
3.1.3. The comprehensive model rank (M.R)
Table 3 provides the ranking for all models analyzed using 8 climate
indices at the annual scale over the five macro regions of Brazil. For NT
Brazil the best performances are found with the CMIP5 and CMIP6
models. The three best models are MIROC5 (20), CNRM-ESM2-1 (28)
and CNRM-CM6-1-HR (27) with Mr values of 0.874, 0.863 and 0.842,
respectively (Table 3). In Northeast Brazil (NE), the three best models
belong to distinct generations of CMIP. The best model (FGOALS-G2) is
9
Weather and Climate Extremes 38 (2022) 100511
from CMIP5, the second one (MPI-ECHAM5) of CMIP3, and the third
(MPI-ESM-1-2-LR) belongs to CMIP6. However, the five-worst models
are from CMIP3. Indeed, the MME shows that the CMIP3 models
represent more accurately the extreme precipitation climate indices in
NE. This indicates that all the improvements made over the years in the
ESMs did not imply in better representation of precipitation climate
extremes over NEB, even though this region is one of the most vulner­
able to extreme events in South America (Torres et al., 2012a; Dare­
la-Filho et al., 2016; Marengo et al., 2020b; Ribeiro et al., 2021;
Zachariah et al., 2022).
In the Center-West (CW) the relative best performance is found again
with the CMIP5 and CMIP6 models. However, unlike previously
observed in the NT region, CMIP5 models’ performance is slightly better
than CMIP6. The three best models - MPI-ESM-1-2-HR (38), MPI-ESM-
MR (22) e MIROC-ESM-CHEM (18) – show Mr values above 0.85
(Table 3). The best performance for the Southeast (SE) and South (ST)
regions of Brazil is with the CMIP6 models. A special highlight is given
for ST region in which the three best models belong to CMIP6 (MIROC6,
GFDL-ESM4 e CNRM-CM6-1-HR), a pattern not observed in the other
regions of Brazil.
Overall, our results indicate that neither model stands out as the best
for all Brazilian regions. Moreover, only in the NT, SE, and ST regions
the evolution of ESMs shows improvements in the representation of
precipitation climate extremes, considering that in these regions the
results with the CMIP6 models were better concerning CMIP3 and
CMIP5. Therefore, our results suggest that, at least for Brazil, the evo­
lution of the ESMs from CMIP apparently did not reflect a substantial
improvement in the representation of precipitation climate extremes for
all Brazilian regions. However, further studies using other methods and
variables (e.g., solar radiation and temperature) are encouraged to
analyze the improvement of CMIP generations. Based on this result, we
can infer that despite the dynamical core of models working well for
Brazil, the physics of the CMIP6 models still needs improvements to
represent more accurately some climatic aspects worldwide (e.g., pre­
cipitation climate extremes).
3.2. Trend analysis
Analyzing the observed climate trends considering the average area
F. Jeferson de Medeiros et al. Weather and Climate Extremes 38 (2022) 100511

Table 3
Regional comprehensive ranking (Rank) and skill score (MR ) for extreme precipitation climate indices at the annual scale for individual CMIP3, CMIP5 and CMIP6
models, and their multi-model ensembles (MME) over the five Brazilian macro regions. The best (worst) three models for each Brazilian region are highlighted in bold
(italics).
Models NT NE CW SE ST

Rank MR Rank MR Rank MR Rank MR Rank MR

CMIP3 01 CNRM-CM3 25 0,733 17 0,757 16 0,761 23 0,741 38 0,647

02 FGOALS-1-0-G 42 0,602 26 0,726 32 0,706 32 0,708 39 0,632
03 GFDL-CM2-0 28 0,719 14 0,776 29 0,726 5 0,829 34 0,660
04 GFDL-CM2-1 33 0,689 10 0,806 14 0,770 31 0,710 40 0,624
05 IPSL-CM4 43 0,597 29 0,717 41 0,624 37 0,679 42 0,615
06 MIROC3-2-HIRES 6 0,824 13 0,777 15 0,765 4 0,841 9 0,818
07 MIROC3-2-MEDRES 21 0,739 28 0,720 24 0,745 22 0,747 29 0,703
08 MPI-ECHAM5 14 0,787 2 0,883 9 0,787 11 0,782 4 0,869
09 MRI-CGMC2-3-2A 27 0,727 16 0,764 28 0,735 35 0,706 32 0,663
CMIP5 10 CNRM-CM5 34 0,682 32 0,693 17 0,758 19 0,753 14 0,794
11 FGOALS-G2 19 0,745 1 0,891 34 0,685 27 0,730 31 0,665
12 FGOALS-S2 37 0,673 23 0,739 31 0,715 30 0,715 43 0,578
13 GFDL-CM3 23 0,737 35 0,674 30 0,721 16 0,765 35 0,657
14 IPSL-CM5A-LR 38 0,671 11 0,779 32 0,706 40 0,646 41 0,616
15 IPSL-CM5A-MR 26 0,730 8 0,821 26 0,737 42 0,598 37 0,649
16 IPSL-CM5B-LR 40 0,661 31 0,695 34 0,685 26 0,734 27 0,710
17 MIROC-ESM-CMIP5 31 0,705 6 0,830 18 0,756 23 0,741 23 0,754
18 MIROC-ESM-CHEM 41 0,653 7 0,823 3 0,859 9 0,797 36 0,650
19 MIROC4H 17 0,756 27 0,721 38 0,674 33 0,707 18 0,772
20 MIROC5 1 0,874 21 0,747 43 0,604 43 0,593 5 0,855
21 MPI-ESM-LR 4 0,833 5 0,846 12 0,779 28 0,726 13 0,802
22 MPI-ESM-MR 13 0,790 32 0,693 2 0,869 6 0,808 10 0,816
23 MPI-ESM-P 8 0,819 30 0,715 6 0,817 18 0,760 7 0,848
24 MRI-CGMC3 35 0,679 43 0,595 20 0,752 8 0,799 16 0,773
25 MRI-ESM1 10 0,807 35 0,674 5 0,840 2 0,880 21 0,763
CMIP6 26 CNRM-CM6 12 0,798 37 0,671 11 0,780 1 0,892 11 0,812
27 CNRM-CM6-1-HR 3 0,842 39 0,664 22 0,750 13 0,777 3 0,876
28 CNRM-ESM2-1 2 0,863 41 0,644 13 0,777 14 0,775 26 0,712
29 FGOALS-F3-L 32 0,698 4 0,850 40 0,640 33 0,707 8 0,822
30 FGOALS-G3 20 0,743 18 0,750 4 0,842 9 0,797 6 0,849
31 GFDL-ESM4 18 0,755 42 0,624 39 0,651 29 0,721 2 0,909
32 IPSL-CM5A2-INCA 30 0,707 9 0,819 20 0,752 39 0,672 33 0,661
33 IPSL-CM6A-LR 39 0,669 40 0,658 37 0,676 15 0,766 24 0,733
34 IPSL-CM6-LR-INCA 36 0,677 38 0,665 36 0,683 36 0,687 22 0,760
35 MIROC-ES2L 23 0,737 24 0,728 8 0,791 38 0,674 28 0,708
36 MIROC6 5 0,829 34 0,675 42 0,620 41 0,609 1 0,938
37 MPI-ESM-1-2-HAM 16 0,764 18 0,750 18 0,756 17 0,762 20 0,766
38 MPI-ESM-1-2-HR 11 0,801 12 0,778 1 0,870 7 0,806 18 0,772
39 MPI-ESM-1-2-LR 9 0,815 3 0,880 10 0,782 12 0,779 12 0,804
40 MRI-ESM2-0 7 0,823 20 0,749 7 0,798 3 0,864 16 0,773
41 MME-CMIP3 29 0,713 15 0,770 27 0,735 21 0,749 30 0,670
42 MME-CMIP5 22 0,738 22 0,746 23 0,747 25 0,735 25 0,710
43 MME-CMIP6 15 0,768 25 0,727 25 0,745 20 0,753 15 0,799

The optimal value of MR is 1.0.

for each Brazilian region (Fig. 8) as well as its spatial distribution
(Fig. 9), it can be seen that for the NT region negative trends are pre­
dominant (decrease) for all precipitation indices, except for CWD
(Fig. 8). An inverse pattern is found for NE Brazil; that is, with the
exception of CWD, all the remaining precipitation indices show a posi­
tive trend. All the observed trends for the Center-West and Southeast
regions are negative (Fig. 8). In the South region, 4 precipitation indices
exhibit negative trends (PRCPTOT, RX1day, RX5day, and CWD) and 4
positives (R95p, SDII, R20mm, and CDD). The model performance varies
according to the region and precipitation climate index analyzed.
However, in general, the individual model performance and its multi-
model ensemble have difficulty in simulating the observed trends, as
also observed in Fig. 9. A similar result was found by Avila-Diaz et al.
(2020b) for the eight hydrological basins of Brazil. The most consider­
able inconsistencies are present in RX1day (Fig. 8c) and RX5day
(Fig. 8d). For example, the intensity of the observed trends for RX1day
are lower than − 0.8 mm decade− 1 for the North, Center-West, South­
east, and South regions of Brazil. In contrast, the simulated trends
depicted by different model ensemble shows a predominance of positive
trends (>0.2 mm decade− 1). For other indices, such as R95p (Fig. 8b)
and R20mm (Fig. 8f) the multi-model ensembles correctly represent the
observed trend signal in most regions; however, they underestimate
their intensity, especially in Northern and Southeastern Brazil. Over
these two regions the observed trends are − 0.5 and − 1.2 mm decade− 1
for R95p, and − 1.0 days decade− 1 for R20mm, while the simulated
trends depicted by the different model ensembles are − 0.2 mm deca­
de− 1, and -0.1 days decade− 1 for R95p and R20mm, respectively.
Examining the observed trends spatially (Fig. 9 – first column), a
decreasing trend are observed mainly in the region modulated by the
South American monsoon system (Jones and Carvalho, 2013), though
this trend is not statistically significant. In addition, we highlight a
decreasing trend observed in the western portion of Northern Brazil in
most of the precipitation climate indices and a decrease in the number of
consecutive dry days over the Center-West, a region in which the agri­
cultural and livestock have been increasing in the last decades (Battisti
and Sentelhas, 2019; Reis et al., 2020). Although these remarkable
trends over these areas, the models fail to satisfactorily represent the
spatial trends and their intensity, especially with the CMIP5 and CMIP6
models. We hypothesize that unrealistic land-atmosphere interactions
and misrepresentation of the Amazon evapotranspiration (Baker et al.,

10
F. Jeferson de Medeiros et al. Weather and Climate Extremes 38 (2022) 100511

Fig. 8. Decadal trends from 1981 to 2000 in each extreme precipitation climate indices at the annual scale for individual CMIP3, CMIP5 and CMIP6 models, their
multi-model ensemble (MME) and reference database (OBS) over the five Brazilian macro regions. Diagonal lines indicate significant trends at 95% level.

11
F. Jeferson de Medeiros et al. Weather and Climate Extremes 38 (2022) 100511

Fig. 9. Spatial distribution of decadal trends during 1981–2000 for each extreme precipitation climate index at the annual scale over Brazil from observation and the
multi-model ensemble of CMIP3, CMIP5 and CMIP6. Cross-hatching regions indicate a significant trend at 95% level.

12
F. Jeferson de Medeiros et al. Weather and Climate Extremes 38 (2022) 100511

2021) could be related to the spatial inconsistencies. The same is
observed for the CWD in the North region, in which the observation
points out a positive trend while the multi-model ensembles negative
trends. It is worth mentioning that CMIP6 almost generalized trends
towards an increase in CDD in Brazil. Regarding the statistical signifi­
cance, the extreme precipitation climate indices show significant trends
only in small areas of Brazil, which is in line with those found by Skansi
et al. (2013), Avila-Diaz et al. (2020b) and Reboita et al. (2022).
As the analyses indicated that the skill of the models varies according
to the region, there is not a single model that is the most appropriate to
represent the observed trends, and the evolution of ESMs from CMIP did
not reflect a substantial improvement in the representation of precipi­
tation climate extremes over all Brazilian regions, the climate pro­
jections will be performed considering the multi-model ensemble
(MMEs) of each CMIP generation (CMIP3, CMIP5 and CMIP6). This
technique has been widely used to reduce uncertainties and increase

Fig. 10. Projected changes (%) in the extreme precipitation climate index of PRCPTOT (a–f) and R95p (g–l) with the multi-model ensemble (MME) of CMIP3, CMIP5
and CMIP6 for the middle (2046–2065) and far future (2081–2 100) under its worst-case climate scenario. The changes are relative to 1981–2000 period.

13
F. Jeferson de Medeiros et al.
climate projection reliability (Tebaldi and Knutti, 2007; Torres and
Marengo, 2013; Martre et al., 2015; Reboita et al., 2022; Gouveia et al.,
2022).
3.3. Future changes in precipitation climate extremes
The climate projections for the eight extreme precipitation climate
indices with the multi-model ensemble of CMIP3, CMIP5 and CMIP6
under its worst-case climate scenario for the middle (2046–2065) and
far future (2081–2 100) are shown in Figs. 10–13. For the PRCPTOT
Fig. 11. As in Fig. 10, but for RX1day (a–f) and RX5day (g–l).
14
Weather and Climate Extremes 38 (2022) 100511
index (Fig. 10a–f), the signal of projected changes depends on the gen­
eration of CMIP considered, except for South Brazil, where an increase is
expected in all cases. Brazil’s northern region is expected to increase
with CMIP3, especially in the western portion (Fig. 10a, d). However,
the CMIP5 and CMIP6 models project a decrease, mainly in CMIP6.
These changes follow the negative simulated trends in the present
climate, though this trend is not statistically significant (Fig, 8a). The
largest decrease is expected toward the end of the 21st century and in
the northern area, where the rainfall is mainly modulated by the squall
lines and ITCZ meteorological systems (Cohen et al., 1995; Alcântara
F. Jeferson de Medeiros et al.
Fig. 12. As in Fig. 10, but for SDII (a–f) and R20mm (g–l).
et al., 2011; Medeiros et al., 2020b; Luiz-Silva et al., 2021; Espinoza
et al., 2022). On average, the climate projections indicate an increase of
3.9% with the MME of CMIP3 and a decrease of − 1.5% and − 10.0% in
CMIP5 and CMIP6, respectively (Table 4). Therefore, the observed
negative trends in the present climate combined with the projected
decrease of PRCPTOT, especially by the MME of CMIP6, indicate that
the reduction in PRCPTOT may be intensified in the future, which would
impact various socioeconomic sectors, such as agriculture and water
resources (Medeiros et al., 2022). In Northeast Brazil, the projections are
also divergent across CMIP generations. In this case, we highlight the
15
Weather and Climate Extremes 38 (2022) 100511
projected changes by the CMIP5 models (Fig. 11b and e), which project
an increase of PRCPTOT over the entire region, while in CMIP3 and
CMIP6 the signal varies depending on the area analyzed, with the
southern (northern) portion being projected a decrease (increase). The
projected increase over Northern Northeast Brazil may be related to a
potential wider displacement of the ITCZ, as shown by Arteaga and Lima
(2020). In percentage terms, the projected changes at the end of the
century with the MME of CMIP3, CMIP5 and CMIP6 are 0.5%, 20.3%
and − 2.4%, respectively (Table 4).
For the Center-West and Southeast regions, the signal of the
F. Jeferson de Medeiros et al.
Fig. 13. As in Fig. 10, but for CDD (a–f) and CWD (g–l).
projected changes also depends on the generation of CMIP analyzed.
However, unlike was found in the North and Northeast regions, it is
expected that PRCPTOT will decrease in CMIP3, increase in CMIP5, and
again decrease with CMIP6 (Fig. 10a–f). This indicates that depending
on the version of ESMs from CMIP used, the climate projections of
PRCPTOT may be reasonably different. Despite this divergence across
CMIP generations, it is important to highlight that at the end of the 21st
century (2081–2 100) the projected magnitudes changes are more pro­
nounced than in the middle future (2046–2065), in agreement with is
16
Weather and Climate Extremes 38 (2022) 100511
usually found in the literature (e.g. Almazroui et al., 2021; Lovino et al.,
2021). The projected changes for these regions may exceed 10%
(Table 4).
The climate projections for R95p (Fig. 10g-l) indicate an increase in
most regions of Brazil and across all CMIP generations (Table 4). The
increase may exceed 10% for the South region, being 17.2% in CMIP6
(Table 4). The R95p index is associated with the occurrence of flash
floods (Tebaldi et al., 2006; Blázquez and Solman, 2020; Funatsu et al.,
2021; Faye and Akinsanola, 2022). Thus, their positive projected
F. Jeferson de Medeiros et al.
Table 4
Average projected changes for each extreme precipitation climate indices over the five Brazilian macro regions
with the multi-model ensemble of CMIP3, CMIP5 and CMIP6 for the far future (2081–2 100) under its worst-
case climate scenario. Cells in blue (brown) indicate changes to a wetter (drier) climate. Values in bold
represent the highest change for each index in each CMIP generation. The changes are relative to 1981–2000
period.
changes elevate the risk of flash floods over the Brazilian territory and
indicate that these events are likely to be more severe in the future.
These results are consistent with that of Sillmann et al. (2013b), Avi­
la-Diaz et al. (2020a), and Gouveia et al. (2022), who reported an in­
crease in the intensity of very wet days. Furthermore, the expected
increase of R95p combined with the changes in PRCPTOT suggests that
changes in rainfall distribution may occur, with an expected increase in
the number of dry days, as described by Medeiros and Oliveira (2022)
for Northeast Brazil.
For the maximum 1-day (RX1day) and consecutive 5-day precipita­
tion (RX5day) the projections for both the middle and far future indicate
an increase of these extreme events over all Brazilian regions and CMIP
generations (Fig. 11), which is in line with previous findings (Bador
et al., 2018; Reboita et al., 2022). In this case, the greatest magnitude
changes are expected with the MME of CMIP3 (Fig. 11d and j), especially
for the North and Northeast region of Brazil, in which are expected an
increase of 94.3% and 79.0% for RX1day and 59.1% and 54.0% for
RX5day, respectively (Table 4). It is important to highlight that in
Northeast Brazil the CMIP3 models showed the best performance in the
historical climate (Fig. 4 and Table 3), which indicates greater reliability
for the projections with this CMIP generation, despite all the un­
certainties associated with the climate projections (Knutti and Sedlácek,
2013; Lehner et al., 2020). On the other hand, the best performance in
the historical climate for the North region was obtained with the CMIP6
models (Fig. 3 and Table 3). Within this context, although the climate
projections still indicate an increase in extreme events, they are less
severe (33.6% for RX1day and 20.3% for RX5day) compared to the
projections with the CMIP3 (Table 4). The same is observed in the
Southeast and South Brazil regions, in which the projected changes at
the end of the century with the CMIP6 are 31.8% and 42.5% for RX1day
(Fig. 11f and Table 4) and 27.8% and 28.9% for RX5day (Fig. 11l and
Table 4), respectively.
Extreme flash floods and heavy rainfall events are among Brazil’s
most common hydrological hazards (CEPED, 2013; Ribeiro et al., 2021;
Aguiar and Cataldi, 2021). Thus, the projected increase of RX1day and
RX5day indicate that device policies of adaptation and mitigation must
be developed to reduce the probably associated impacts (e.g., land­
slides) that future changes in these extreme events may cause over
Brazil, as also warned by Zachariah et al. (2022).
Future changes for SDII (Fig. 12a–f) and R20mm (Fig. 12g-l) show
similar behavior to that described for RX1day and RX5day (Fig. 11). The
SDII and R20mm are projected to increase throughout the 21st century,
17
Weather and Climate Extremes 38 (2022) 100511
with the largest relative change in the NT and NE regions, especially for
R20mm with the MME of CMIP3 (Fig. 12g, j) and CMIP5 (Fig. 12h, k).
The climate projections show that the SDII and R20mm extreme pre­
cipitation indices are expected to be less severe and frequent with the
evolution of ESMs from CMIP. This indicates that the CMIP6 models may
project the precipitation climate extremes with less uncertainty. More­
over, it also suggests that the CMIP6 models may be less sensitive in
representing the precipitation climate indices associated with intensity
and frequency of rainfall, although some studies have shown that CMIP6
models project greater warming towards the end of the century relative
to CMIP5 and CMIP3 (Zelinka et al., 2020; Dong et al., 2020; IPCC et al.,
2021).
In percentage terms (Table 4), the SDII are expected to increase
8.3%, 11.5% and 18.7% over the NT, SE and ST regions, respectively, at
the end of the century with the MME of CMIP6. For R20mm, an increase
of 36.8% and 168.0% are projected over CW and NE with the CMIP5 and
CMIP3, respectively. The increase in the intensity of SDII combined with
the projected changes of the PRCPTOT (Fig. 10a–f) suggest a future
concentration of total precipitation in fewer rainfall events or the
occurrence of heavy rainfall will be more severe in the future. These
findings are consistent with those previously reported by Marengo et al.
(2020b) and Medeiros and Oliveira (2022).
Finally, the number of CDD and CWD are expected to increase and
decrease, respectively, over all Brazilian regions (Fig. 13 and Table 4),
which indicates conditions more favorable for water deficit in the future.
The largest future changes of CDD are projected in Northern Brazil, the
east coast of NE and over the SE and ST regions, especially with the
CMIP6 (Fig. 13f). In percentage terms, the projected magnitude changes
reach 49.7%, 28.4%, 20.5%, 28.7%, and 18.7% in the NT, NE, CW, SE
and ST regions, respectively (Table 4). For CWD, the largest increase
(more than 20%) is expected over the NT region in all CMIP generations.
The climate change signal and magnitude could also be improved and
amplified if downscaling procedures and bias correction methods had
been applied, as shown by Avila-Diaz et al. (2020a) and Iturbide et al.
(2022). This feature will be the subject of future works by the authors.
The summary of the main projected changes expected for precipi­
tation climate extremes over each Brazilian region is depicted in Fig. 14.
Several papers in the literature also point to projections similar to those
found in our study (Avila-Diaz et al., 2020a; Gouveia et al., 2022;
Reboita et al., 2022). These drier projected conditions highlight that the
rainfall scarcity usually observed in NEB must be aggravated in the
future due to climate change. Moreover, it is an indication that the water
F. Jeferson de Medeiros et al.
Fig. 14. Schematic summary of the main projected changes in precipitation
climate extremes over the Brazilian regions.
deficit and seasonal volume of rainfall may reduce, as already described
by previous studies (Oliveira et al., 2017; Marengo et al., 2020b;
Medeiros et al., 2022; Medeiros and Oliveira, 2022). The increased
projected changes of CWD combined with the decrease in CDD may also
fuel additional dry spells and enhance the risk of forest fires (Aragão
et al., 2007). In this case, we highlight the Brazilian Pantanal wetlands
that recently have been showing an increasing trend in these extreme
events (Libonati et al., 2020; Marengo et al., 2021). It is also important
to highlight the projected magnitude increase of CDD in the south­
eastern portion of NE in all CMIP generations. This area has been
recently considered a new Brazilian agricultural frontier (Silva et al.,
2022).
4. Conclusions
Understanding whether the evolution of ESMs from CMIP improves
the representation of the current climate is of great importance because
device policies of adaptation and mitigation to climate change generally
consider the results of the most recent generation of ESMs. Moreover,
comprehend how the climate models simulate the observed climate
extremes provides useful information about its reliability in investi­
gating future changes. In this sense, we assessed the performance of a
subset of 40 ESMs from CMIP3, CMIP5, and CMIP6 in representing
extreme precipitation climate indices over Brazil during 1981–2000.
Additionally, we also estimated their projected changes for the middle
(2046–2065) and far future (2081–2 100) under the worst-case scenario
for each CMIP generation.
In the historical period, the results showed that the performance of
the ESMs depends on the extreme precipitation index investigated and
the macro region of Brazil analyzed. The North and Northeast (Southeast
and South) regions show the greatest (lowest) uncertainties in the rep­
resentation of climate extremes. This highlighted that the ESMs show
deficiencies in representing precipitation climate extremes in tropical
South America, indicating that continuous improvements are needed in
the CMIP models. The RX1day, CDD, and CWD ETCCDI indices were the
worst simulated over Brazil, while the best ones were PRCPTOT, R20mm
and SDII, in agreement with found by Avila-Diaz et al. (2020b) with
ESMs at high horizontal resolution.
Regarding the individual model’s performance for each CMIP gen­
eration, our results showed that, in general, neither model stands out as
18
Weather and Climate Extremes 38 (2022) 100511
the best, and all of them have difficulty simulating trends. However, the
multi-model ensemble indicated that the CMIP3 models represent more
precisely the precipitation extremes in Northeast Brazil. For Center-
West, the best performance was found for CMIP5. On the other hand,
in the North, Southeast, and South regions of Brazil the CMIP6 models
showed the best relative performance (Fig. 14). Thus, our results suggest
that CMIP6 appears to have advanced in the simulation of precipitation
climate extremes in mid-latitudes. However, at least for Brazil, the
evolution of the ESMs from CMIP did not reflect a substantial
improvement in the representation of precipitation climate extremes
over all Brazilian regions.
The projections with the three generations of ESMs are consensus in
signaling an intensification of precipitation extremes over Brazil, which
increases the reliability of the climate projections. The only exception is
for PRCPTOT, in which the projected changes may increase or decrease
depending on the CMIP generation analyzed (Table 4). The summary of
the main projected changes expected for precipitation climate extremes
is depicted in Fig. 14. In percentage terms, the annual maximum 1-day
(RX1day) and consecutive 5-day precipitation (RX1day and RX5day) are
expected to increase by more than 20%. The number of very heavy
precipitation days (20 mm) is projected to increase mainly over North­
east Brazil, with changes reaching 168.0%, 106.7%, and 36.9% in the
CMIP3, CMIP5, and CMIP6, respectively. The simple daily intensity
(SDII) is expected to increase between 8.3% and 24.8%, while the CDD
are projected to increase between 11.2% and 49.7%. On the other hand,
the consecutive wet days (CWD) may decrease up to − 28.9% in North
Brazil region in the far future.
The precipitation extremes are expected to become more severe,
frequent, and long-lasting in Brazil due to climate change. This highlight
that extreme events of flash floods, landslides, and heavy rainfall can hit
with great intensity and more often the Brazilian regions. Likewise,
episodes of rain scarcity must be exacerbated in the future, as the pro­
jections of dry spells indicate that they will be long-lasting. Therefore, as
Brazil is strongly vulnerable to extreme events (CEPED, 2013), adequate
adaptation and mitigation strategies related to climate change impacts
must be reinforced and developed in the following years. Lastly, it is
important to highlight that this vulnerability must be even greater if the
projected increase of heat-waves and extreme compound events in a
warming world is considered (AghaKouchak et al. 2020; Libonati et al.,
2022).
Authors contributions
Felipe Jeferson de Medeiros: Conceptualization; Data curation;
Formal analysis; Methodology; Software; Validation; Visualization;
Writing – original draft; Writing – review & editing. Cristiano Prestrelo
de Oliveira: Conceptualization; Supervision; Writing – review & edit­
ing; Funding acquisition. Álvaro Ávila-Diaz: Data curation; Methodol­
ogy; Software; Writing – review & editing.
Declaration of competing interest
The authors declare that they have no known competing financial
interests or personal relationships that could have appeared to influence
the work reported in this paper.
Data availability
Data will be made available on request.
Acknowledgements
This study was financed in part by Coordenação de Aperfeiçoamento
de Pessoa de Nível Superior – Brasil (CAPES) – Finance Code 001. We
thank the climate modeling groups for producing and making available
their model output, the Earth System Grid Federation (ESGF) for
F. Jeferson de Medeiros et al.
archiving the data and providing access, and the multiple funding
agencies who support CMIP3, CMIP5, CMIP6 and ESGF. The authors
thank the anonymous reviewers for their constructive comments and
suggestions.
References
Aerenson, T., Tebaldi, C., Sanderson, B., Lamarque, J.-F., 2018. Changes in a suite of
indicators of extreme temperature and precipitation under 1.5 and 2 degrees
warming. Environ. Res. Lett. 13, 035009 https://doi.org/10.1088/1748-9326/
aaafd6.
Agel, L., Barlow, M., 2020. How well do CMIP6 models runs match observed Northeast
US precipitation and extreme precipitation-related circulation? J. Clim. 33,
9835–9848. https://doi.org/10.1175/JCLI-D-19-1025.1.
AghaKouchak, A., Chiang, F., Huning, L.S., Love, C.A., Mallakpour, I., Mazdiyasni, O.,
Moftakhari, H., Papalexiou, S.M., Ragno, E., Sadegh, M., 2020. Climate extremes and
compound hazards in a warming world. Annu. Rev. Earth Planet Sci. 48, 519–548.
https://doi.org/10.1146/annurev-earth-071719-055228.
Aguiar, L.F., Cataldi, M., 2021. Social and environmental vulnerability in Southeast
Brazil associated with the south atlantic convergence zone. Nat. Hazards 109,
2423–2437. https://doi.org/10.1007/s11069-021-04926-z.
Akinsanola, A.A., Kooperman, G.J., Pendergrass, A.G., Hannah, W.M., Reed, K.A., 2020.
Seasonal representation of extremes precipitation indices over the United States in
CMIP6 models present-day simulations. Environ. Res. Lett. 15, 094003 https://doi.
org/10.1088/1748-9326/ab92c1.
Alcântara, C.R., Silva Dias, M.A.F., Souza, E.P., Cohen, J.C.P., 2011. Verification of the
role of the low level jets in Amazon squall lines. Atmos. Res. 100, 36–44. https://doi.
org/10.1016/j.atmosres.2010.12.023.
Almazroui, M., Ashfaq, M., Islam, M.N., Rashid, I.U., Kamil, S., Abid, M.A., O’Brien, E.,
Ismail, M., Reboita, M.S., Sörensson, A.A., Arias, P.A., Alves, L.M., Tippett, M.K.,
Saeed, S., Haarsma, R., Doblas-Reyes, F.J., Saeed, F., Kucharski, F., Nadeem, I., Silva-
Vidal, Y., Rivera, J.A., Ehsan, M.A., Martínez-Castro, D., Muñoz, Á.G., Ali, M.A.,
Coppola, E., Sylla, M.B., 2021. Assessment of CMIP6 performance and projected
temperature and precipitation changes over South America. Earth Syst. Environ. 5,
155–183. https://doi.org/10.1007/s41748-021-00233-6.
Alvalá, R.C.S., Cunha, A.P.M.A., Brito, S.S.B., Seluchi, M.E., Marengo, J.A., Moraes, O.L.
L., Carvalho, M.A., 2017. Drought monitoring in the Brazilian Semiarid region. An.
Acad. Bras. Ciênc. 91, e20170209 https://doi.org/10.1590/0001-
3765201720170209.
Alves, L.M., Chadwick, R., Moise, A., Brown, J., Marengo, J.A., 2021. Assessment of
rainfall variability and future change in Brazil across multiple timescales. Int. J.
Climatol. 41, e1875–e1888. https://doi.org/10.1002/joc.6818.
Aragão, L.E., Malhi, Y., Roman-Cuesta, R.M., Saatchi, S., Anderson, L.O.,
Shimabukuro, Y.E., 2007. Spatial patterns and fire response of recent Amazonian
droughts. Geophys. Res. Lett. 34, 1–5. https://doi.org/10.1029/2006GL028946.
Arias, P.A., Ortega, G., Villegas, L.D., Martínez, J.A., 2021. Colombian climatology in
CMIP5/CMIP6 models: persistent biases and improvements. Revista Facultad de
Ingeniería de Antioquia 100, 75–95. https://doi.org/10.17533/udea.
redin.20210525.
Artaxo, P., 2019. Working together for Amazonia. Science 363, 323. https://doi.org/
10.1126/science.aaw6986.
Arteaga, V.L., Lima, C.H.R., 2020. Analysis of CMIP5 simulations of key climate indices
associated with the South America Monsoon system. Int. J. Climatol. 41, 404–422.
https://doi.org/10.1002/joc.6627.
Ávila, A., Justino, F., Wilson, A., Browich, D., Amorim, M., 2016. Recent precipitation
trends, flash floods and landslides in southern Brazil. Environ. Res. Lett. 11, 114029.
https://doi:10.1088/1748-9326/11/11/114029.
Avila-Diaz, A., Benezoli, V., Justino, F., Torres, R., Wilson, A., 2020a. Assessing current
and future trends of climate extremes across Brazil based on reanalyses and earth
system model projections. Clim. Dynam. 55, 1403–1426. https://doi.org/10.1007/
s00382-020-05272-9.
Avila-Diaz, A., Abrahão, G., Justino, F., Torres, R., Wilson, A., 2020b. Extreme climate
indices in Brazil: evaluation of downscaled earth system models at high horizontal
resolution. Clim. Dynam. 54, 5065–5088. https://doi.org/10.1007/s00382-020-
05333-z.
Ayugi, B., Jiang, Z., Iyakaremye, V., Ngoma, H., Babaousmail, H., Onyutha, C., Dike, V.
N., Mumo, R., Ongoma, V., 2022. East African population exposure to precipitation
extremes under 1.5 C and 2.0 C warming levels based on CMIP6 models. Environ.
◦ ◦
Res. Lett. 17, 044051 https://doi.org/10.1088/1748-9326/ac5d9d.
Bador, M., Donat, M., Geofroy, O., Alexander, L., 2018. Assessing the robustness of future
extreme precipitation intensification in the CMIP5 ensemble. J. Clim. 31,
6505–6525. https://doi.org/10.1175/JCLI-D-17-0683.1.
Bador, M., Boé, J., Terray, L., Alexander, L.V., Baker, A., Bellucci, A., Haarsma, R.,
Koenigk, T., Moine, M., Lohmann, K., Putrasahan, D.A., Roberts, C., Roberts, M.,
Scoccimarro, E., Schiemann, R., Seddon, J., Senan, R., Valcke, S., Vanniere, B., 2020.
Impact of higher spatial atmospheric resolution on precipitation extremes over land
in global climate models. J. Geophys. Res. Atmos. 125, 1–23. https://doi.org/
10.1029/2019JD032184.
Baker, J.C.A., Garcia-Carreras, L., Buermann, W., Castilho De Souza, D., Marsham, J.H.,
Kubota, P.Y., Gloor, M., Coelho, C.A.S., Spracklen, D.V., 2021. Robust Amazon
precipitation projections in climate models that capture realistic land-atmosphere
interactions. Environ. Res. Lett. 16, 074002 https://doi.org/10.1088/1748-9326/
abfb2e.
19
Weather and Climate Extremes 38 (2022) 100511
Battisti, R., Sentelhas, P.C., 2019. Characterizing Brazilian soybean-growing regions by
water deficit patterns. Field Crop. Res. 240, 95–105. https://doi.org/10.1016/j.
fcr.2019.06.007.
Beck, H., Wood, E., Pan, M., Fisher, C.K., Miralles, D.G., van Dijk, A., McVicar, T.R.,
Adler, R.F., 2019. MSWEP V2 global 3-hourly 0.1◦ precipitation: methodology and
quantitative assessment. Bull. Am. Meteorol. Soc. 100, 473–500. https://doi.org/
10.1175/BAMS-D-17-0138.1.
Blázquez, J., Nuñez, M.N., 2013. Analysis of uncertainties in future climate projections
for South America: comparison of WCRP-CMIP3 and WCRP-CMIP5 models. Clim.
Dynam. 41, 1039–1056. https://doi.org/10.1007/s00382-012-1489-7.
Blázquez, J., Solman, S.A., 2020. Multiscale precipitation variability and extremes over
South America: analysis of future changes from a set of CORDEX regional climate
model simulations. Clim. Dynam. 55, 2089–2106. https://doi.org/10.1007/s00382-
020-05370-8.
Bock, L., Lauer, A., Schlund, M., Barreiro, M., Bellouin, N., Jones, C., Meehl, G.A.,
Predoi, V., Roberts, M.J., Eyring, V., 2020. Quantifying progress across different
CMIP phases with the ESMValTool. J. Geophys. Res. Atmos. 125, e2019JD032321
https://doi.org/10.1029/2019JD032321.
Borges, P.A., Franke, J., Silva, F.D.S., Weiss, H., Bernhofer, C., 2014. Differences between
two climatological periods (2001-2010 vs 1971-2000) and trend analysis of
temperature and precipitation in Central Brazil. Theor. Appl. Climatol. 16, 91–202.
https://doi.org/10.1007/s00704-013-0947-4.
Brito, S.S.B., Cunha, A.P.M.A., Cunningham, C.C., Alvalá, R.C., Marengo, J.A.,
Carvalho, M.A., 2017. Frequency, duration and severity of drought in the Semiarid
Northeast Brazil region. Int. J. Climatol. 38, 517–529. https://doi.org/10.1002/
joc.5225.
Brown, P.J., Bradley, R.S., Keimig, F.T., 2010. Changes in extreme climate indices for the
northeastern United States, 1870-2005. J. Clim. 23, 6555–6572. https://doi.org/
10.1175/2010JCLI3363.1.
Builes-Jaramillo, A., Marwan, N., Poveda, G., Kurths, J., 2018. Nonlinear interactions
between the Amazon river basin and the tropical North Atlantic at interannual
timescales. Clim. Dynam. 50, 2951–2969. https://doi.org/10.1007/s00382-017-
3785-8.
Busch, A., Amorim, S., 2012. A tragédia da região serrana do Rio de Janeiro em 2011:
procurando respostas. Casoteca de Gestão Pública, pp. 1–20.
Carvalho, L.M.V., Jones, C., Liebmann, B., 2002. Extreme precipitation events in
Southeastern America and large-scale convective patterns in the South Atlantic
Convergence zone. J. Clim. 15, 2377–2394. https://doi.org/10.1175/1520-0442
(2002)015<2377:EPEISS>2.0.CO;2.
Carvalho, L.M.V., Jones, C., Liebmann, B., 2004. The South Atlantic Convergence Zone:
intensity, form, persistence, and relationships with intraseasonal to interannual
activity and extreme rainfall. J. Clim. 17, 88–108. https://doi.org/10.1175/1520-
0442(2004)017<0088:TSACZI>2.0.CO;2.
Centro Universitário de Estudos e Pesquisas sobre Desastres (CEPED), 2013. Atlas
Brasileiro de Desastres Naturais 1991 a 2012-Volume Brasil, second ed.
Florianópolis. https://s2id.mi.gov.br/pagin as/atlas #. (Accessed 11 February
2022).
Cerón, W.L., Kayano, M.T., Andreoli, R.V., Avila-Diaz, A., Ayes, I., Freitas, E.D.,
Martins, J.A., Souza, R.A.F., 2021. Recent intensification of extreme precipitation
events in the La Plata basin in southern South America (1981-2018). Atmos. Res.
249, 105299 https://doi.org/10.1016/j.atmosres.2020.105299.
Coelho, C.A.S., Oliveira, C.P., Ambrizzi, T., Reboita, M.S., Carpenedo, C.B., Campos, J.L.
P., Tomaziello, A.C.N., Pampuch, L.A., Custódio, M.S., Dutra, L.M.M., Rocha, R.P.,
Rehbein, A., 2016. The 2014 southeast Brazil austral summer drought: regional scale
mechanisms and teleconnections. Clim. Dynam. 46, 3737–3752. https://doi.org/
10.1007/s00382-015-2800-1.
Cohen, J.C.P., Silva Dias, M.A.F., Nobre, C.A., 1995. Environmental conditions
associated with Amazonian squall lines: a case study. Mon. Weather Rev. 123,
3163–3174. https://doi.org/10.1175/1520-0493(1995)123<3163:ECAWAS>2.0.
CO;2.
Cortez, B.N., Pires, G.F., Avila-Diaz, A., Fonseca, H.P., Oliveira, L.R., 2022.
Nonstationary extreme precipitation in Brazil. Hydrol. Sci. J. https://doi.org/
10.1080/02626667.2022.2075267.
Costa, R.L., Baptista, G.M.M., Gomes, H.B., Silva, F.D.S., Rocha Júnior, R.L., Salvador, M.
A., Herdies, D.L., 2020. Analysis of climate extremes indices over northeast Brazil
from 1961 to 2014. Weather Clim. Extrem. 28, 100254 https://doi.org/10.1016/j.
wace.2020.100254.
Darela-Filho, J., Lapola, D., Torres, R., Lemos, M., 2016. Socio-climatic hotspots in
Brazil: how do changes driven by the new set of IPCC climatic projections affect their
relevance for policy? Clim. Change 136, 413–425. https://doi.org/10.1007/s10584-
016-1635-z.
Debortoli, N.S., Camarinha, P.I.M., Marengo, J.A., Rodrigues, R.R., 2017. An index of
Brazil’s vulnerability to expected increases in natural flash flooding and landslide
disasters in the context of climate change. Nat. Hazards 86, 557–582. https://doi.
org/10.1007/s11069-016-2705-2.
Deser, C., Lehner, F., Rodgers, K.B., Ault, T., Delworth, T.L., Dinezio, P.N., Fiore, A.,
Frankignoul, C., Fyfe, J.C., Horton, D.E., Kay, J.E., Knutti, R., Lovenuski, N.S.,
Marotzke, J., McKinnon, K.A., Randerson, J., Screen, J.A., Simpson, I.R., Ting, M.,
2020. Insights from earth system model initial-condition large ensembles and future
prospects. Nat. Clim. Change 10, 277–286. https://doi.org/10.1038/s41558-020-
0731-2.
Díaz, L.B., Saurral, R.I., Vera, C.S., 2021. Assessment of South America summer rainfall
climatology and trends in a set of global climate models large ensembles. Int. J.
Climatol. 41, e59–e77. https://doi.org/10.1002/joc.6643.
F. Jeferson de Medeiros et al.
Donat, M., Alexander, L., Yang, H., Durre, I., Vose, R., Caesar, J., 2013. Global land-based
datasets for monitoring climatic extremes. Bull. Am. Meteorol. Soc. 94, 997–1006.
https://doi.org/10.1175/BAMS-D-12-00109 .1.
Donat, M.G., Lowry, A.L., Alexander, L.V., O’Gorman, P.A., Maher, N., 2016. More
extreme precipitation in the world’s dry and wet regions. Nat. Clim. Change 6,
508–513. https://doi.org/10.1038/nclimates2941.
Dong, Y., Armour, K.C., Zelinka, M.D., Proistosescu, C., Battisti, D.S., Zhou, C.,
Andrews, T., 2020. Intermodel spread in the patter effect and its contribution to
climate sensitivity in CMIP5 amd CMIP6 models. J. Clim. 33, 7755–7775. https://
doi.org/10.1175/JCLI-D-19-1011.1.
Drumond, A., Marengo, J., Ambrizzi, T., Nieto, R., Moreira, L., Gimeno, L., 2014. The
role of the Amazon basin moisture in the atmospheric branch of the hydrological
cycle: a Lagrangian analysis. Hydrol. Earth Syst. Sci. 18, 2577–2598. https://doi.
org/10.5194/hess-18-2577-2014.
Durkee, J.D., Mote, T.L., Shepherd, J.M., 2009. The contribution of mesoscale convective
complexes to rainfall across subtropical South America. J. Clim. 22, 4590–4605.
https://doi.org/10.1175/2009JCLI2858.1.
Espinoza, J.C., Ronchail, J., Frappart, F., Lavado, W., Santini, W., Guyot, J.L., 2013. The
major floods in the Amazonas river and tributaries (Western Amazon basin) during
the 1970–2012 period: a focus on the 2012 flood. J. Hydrometeorol. 14, 1000–1008.
https://doi.org/10.1175/JHM-D-12-0100.1.
Espinoza, J.C., Marengo, J.A., Schongart, J., Jimenez, J.C., 2022. The new historical
flood in the Amazon River compared to major floods of the 21st century:
atmospheric features in the context of the intensification of floods. Weather Clim.
Extrem. 35, 100406 https://doi.org/10.1016/j.wace.2021.100406.
Eyring, V., Bony, S., Meehl, G.A., Senior, C.A., Stevens, B., Stouffer, R.J., Taylor, K.E.,
2016. Overview of the coupled model intercomparison project Phase 6 (CMIP6)
experimental design and organization. Geosci. Model Dev. (GMD) 9, 1937–1958.
https://doi.org/10.5194/gmd-9-1937-2016.
Faye, A., Akinsanola, A.A., 2022. Evaluation of extreme precipitation indices over West
Africa in CMIP6 models. Clim. Dynam. 58, 925–939. https://doi.org/10.1007/
s00382-021-05942-2.
Funatsu, B.M., Roux, R.L., Arvor, D., Espinoza, J.C., Claud, C., Ronchail, J., Michot, V.,
Dubreuil, V., 2021. Assessing precipitation extremes (1981-2018) and deep
convective activity (2002-2018) in the Amazon region with CHIRPS and AMSU data.
Clim. Dynam. 57, 827–849. https://doi.org/10.1007/s00382-021-05742-8.
Funk, C., Peterson, P., Landsfeld, M., Pedreros, D., Verdin, J., Shukla, S., Husak, G.,
Rowland, J., Harrison, L., Hoell, A., Michaelsen, J., 2015. The climate hazards
infrared precipitation with stations—a new environmental record for monitoring
extremes. Sci. Data 2, 150066. https://doi.org/10.1038/sdata.2015.66.
Gomes, H.B., Ambrizzi, T., Silva, B.F.P., Hodges, K., Silva Dias, P.L., Herdies, D.L.,
Silva, M.C.L., Gomes, H.B., 2019. Climatology of easterly wave disturbances on the
tropical South Atlantic. Clim. Dynam. 53, 1393–1411. https://doi.org/10.1007/
s00382-019-04667-7.
Gouveia, C.D., Torres, R.R., Marengo, J.A., Avila-Diaz, A., 2022. Uncertainties in
projections of climate extremes in South America via Bayesian inferecence. Int. J.
Climatol. https://doi.org/10.1002/joc.7650.
Grimm, A.M., Ferraz, S.E.T., Gomes, J., 1998. Precipitation anomalies in southern Brazil
associated with el niño and La niña events. J. Clim. 11, 2863–2880. https://doi.org/
10.1175/1520-0442(1998)011<2863:PAISBA>2.0.CO:2.
Gulizia, C., Camilloni, I., 2015. Comparative analysis of the ability of a set of CMIP3 and
CMIP5 global climate models to represent precipitation in South America. Int. J.
Climatol. 35, 583–595. https://doi.org/10.1002/joc.4005.
Gusain, A., Ghosh, S., Karmakar, S., 2020. Added value of CMIP6 over CMIP5 models in
simulating Indian summer monsoon rainfall. Atmos. Res. 232, 104680 https://doi.
org/10.1016/j.atmosres.2019.104680.
Hasler, N., Avissar, R., 2007. What controls evapotranspiration in the Amazon Basin?
J. Hydrometeorol. 8, 380–395. https://doi.org/10.1175/JHM587.1.
Huffman, G.J., Adler, R.F., Morrissey, M.M., Bolvin, D.T., Curtis, S., Joyce, R.,
McGavock, B., Susskind, J., 2001. Global precipitation at one-degree daily resolution
from multisatellite observations. J. Hydrol. 2, 36–50. https://doi.org/10.1175/
1525-7541(2001)002<0036:GPAODD>2.0.CO;2.
IPCC, 2018. In: Masson-Delmotte, V., Zhai, P., Pörtner, H.-O., Roberts, D., Skea, J.,
Shukla, P.R., Pirani, A., Moufouma-Okia, W., Péan, C., Pidcock, R., Connors, S.,
Matthews, J.B.R., Chen, Y., Zhou, X., Gomis, M.I., Lonnoy, E., Maycock, T.,
Tignor, M., Waterfield, T. (Eds.), Global Warming of 1.5◦ C. An IPCC Special Report
on the Impacts of Global Warming of 1.5◦ C above Pre-industrial Levels and Related
Global Greenhouse Gas Emission Pathways, in the Context of Strengthening the
Global Response to the Threat of Climate Change, Sustainable Development, and
Efforts to Eradicate Poverty (in press).
IPCC, 2021. In: Zhai, V.P., Pirani, A., Connors, S.L., Péan, C., Berger, S., Caud, N.,
Chen, Y., Goldfarb, L., Gomis, M.I., Huang, M., Leitzell, K., Lonnoy, E., Matthews, J.
B.R., Maycock, T.K., Waterfield, T., Yelekçi, O., Yu, R., Zhou, B. (Eds.), Climate
Change 2021: the Physical Science Basis. Contribution of Working Group I to the
Sixth Assessment Report of the Intergovernmental Panel on Climate Change
[Masson-Delmotte. Cambridge University Press (in press).
Iturbide, M., Casanueva, A., Bedia, J., Herrera, S., Milovac, J., Gutiérrez, J.M., 2022. On
the need of bias adjustment for more plausible climate change projections of extreme
heat. Atmos. Sci. Lett. 23, e1072 https://doi.org/10.1002/asl.1072.
Jesus, E.M., Rocha, R.P., Crespo, N.M., Reboita, M.S., Gozzo, L.F., 2021. Multi-model
climate projections of the main cyclogenesis hot-spots and associated winds over the
eastern coast of South America. Clim. Dynam. 56, 537–557. https://doi.org/
10.1007/s00382-020-05490-1.
Jiang, Z., Li, W., Xu, J., Li, L., 2015. Extreme precipitation indices over China in CMIP5
models. Part I: model evaluation. J. Clim. 28, 8603–8619. https://doi.org/10.1175/
JCLI-D-15-0099.1.
20
Weather and Climate Extremes 38 (2022) 100511
Jiménez-Muñoz, J.C., Mattar, C., Barichivich, J., Santamaría-Artigas, A., Takahashi, K.,
Malhi, Y., Sobrinho, J.A., Schrier, G.V., 2016. Record-breaking warming and
extreme drought in the Amazon rainforest during the course of El Niño 2015-2016.
Sci. Rep. 6, 33130. https://doi:10.1038/srep33130.
Jones, P.W., 1999. First- and second-order conservative remapping schemes for grids in
spherical coordinates. Mon. Weather Rev. 127, 2204–2210. https://doi.org/
10.1175/1520-0493(1999)127<2204:FASOCR>2.0.CO;2.
Jones, C., Carvalho, L.M.V., 2013. Climate change in the south American monsoon
system: present climate and CMIP5 projections. J. Clim. 26, 6660–6678. https://doi.
org/10.1175/JCLI-D-12-00412.1, 2013.
Kendall, M.G., 1975. Rank Correlation Methods, fourth ed. Charles Griffn, London.
Kim, Y-H., Min, S-K., Zhang, X., Sillmann, J., Sandstad, M., 2020. Evaluation of the
CMIP6 multi-model ensemble for climate extremes indices. Weather Clim. Extrem.
29, 100269. https://doi.org/10.1016/j.wace.2020.100269.
Knutti, R., Sedlácek, J., 2013. Robustness and uncertainties in the new CMIP5 climate
model projections. Nat. Clim. Change 3, 369–373. https://doi.org/10.1038/
nclimates1716.
Kousky, W.E., 1979. Frontal influences on northeast Brazil. Mon. Weather Rev. 107,
1140–1153. https://doi.org/10.1175/1520-0493(1979)107<1140:FIONB>2.0.CO;
2.
lacovone, M.F., Pántano, V.C., Penalba, O.C., 2020. Consecutive dry and wet days over
South America and their association with enso, in cmip5 simulations. Theor. Appl.
Climatol. 142, 791–804. https://doi.org/10.1007/s00704-020-03324-y.
Ledley, T.S., Sundquist, E.T., Schwartz, S.E., Hall, D.K., Fellows, J.D., Killeen, T.L., 1999.
Climate change and greenhouse gases. Eos 80, 453–458. https://doi.org/10.1029/
99EO00325.
Lehner, F., Deser, C., Maher, N., Marotzke, J., Fischer, E.M., Brunner, L., Knutti, R.,
Hawkins, E., 2020. Partitioning climate projection uncertainty with multiple large
ensembles and CMIP5/6. Earth Syst. Dynam. 11, 491–508. https://doi.org/10.5194/
esd-11-491-2020.
Lewis, S.L., Brando, P.M., Phillips, O.L., van der Heijden, G.M.F., Nepstad, D., 2011. The
2010 Amazon drought. Science 331, 554. https://doi.org/10.1126/
science.1200807.
Libonati, R., DaCamara, C.C., Peres, L.F., Carvalho, L.A.S., Garcia, L.C., 2020. Rescue
Brazil’s burning Pantanal wetlands. Nature 588, 217–219. https://doi.org/10.1038/
d41586-020-03464-1.
Libonati, R., Geirinhas, J.L., Silva, P.S., Russo, A., Rodrigues, J.A., Belém, L.B.C.,
Nogueira, J., Roque, F.O., DaCamara, C.C., Nunes, A.M.B., Marengo, J.A., Trigo, R.
M., 2022. Assessing the role of compound drought and heatwave events on
unprecedented 2020 wildfires in the Pantanal. Environ. Res. Lett. 17, 015005
https://doi.org/10.1088/1748-9326/ac462e.
Lovino, M.A., Müller, O.V., Berbery, E.H., Müller, G.V., 2018. Evaluation of CMIP5
retrospective simulations of temperature and precipitation in northeastern
Argentina. Int. J. Climatol. 38, e1158–e1175. https://doi.org/10.1002/joc.5441.
Lovino, M.A., Pierrestegui, M.J., Müller, O.V., Berbery, E.H., Müller, G.V., Pasten, M.,
2021. Evaluation of historical CMIP6 model simulations and future projections of
temperature and precipitation in Paraguay. Climatic Change 164, 46. https://doi.
org/10.1007/s10584-021-03012-4.
Lucas, E.W.M., Sousa, F.A.S., Silva, F.D.S., Rocha Júnior, R.L., Pinto, D.D.C., Silva, V.P.
R., 2021. Trends in climate extreme indices assessed in the Xingu river basin –
Brazilian Amazon. Weather Clim. Extrem. 31, 100306 https://doi.org/10.1016/j.
wace.2021.100306.
Luiz-Silva, W., Garcia, K.C., 2022. Sustainable future and water resources: a synthesis of
the Brazilian hydroelectricity sector in face of climate change. Sustain. Water
Resour. Manag. 8, 120. https://doi.org/10.1007/s40899-022-00711-3.
Luiz-Silva, W., Oscar-Júnior, A.C., Cavalcanti, I.F.A., Treistman, F., 2021. An overview of
precipitation climatology in Brazil: space-time variability of frequency and intensity
associated with atmospheric systems. Hydrol. Sci. J. 66, 289–308. https://doi.org/
10.1080/02626667.2020.1863969.
Luo, N., Guo, Y., Chou, J., Gao, Z., 2021. Added value of CMIP6 models over CMIP5
models in simulating the climatological precipitation extremes in China. Int. J.
Climatol. 42, 1148–1164. https://doi.org/10.1002/joc.7294.
Mann, H.B., 1945. Nonparametric tests against trend. Econometrica 13, 245–259.
https://doi.org/10.2307/1907187.
Marengo, J.A., Tomasella, J., Alves, L.M., Soares, W.R., Rodriguez, D.A., 2011. The
drought of 2010 in the context of historical droughts in the Amazon region. Geophys.
Res. Lett. 38, L12703. https://doi:10.1029/2011GL047436.
Marengo, J.A., Alves, L.M., Alvala, R.C.S., Cunha, A.P., Brito, S., Moraes, O.L.L., 2017a.
Climatic characteristics of the 2010-2016 drought in the semiarid Northeast Brazil
region. An. Acad. Bras. Ciênc 90, 1973–1985. https://doi.org/10.1590/0001-
3765201720170206.
Marengo, J.A., Torres, R.R., Alves, L.M., 2017b. Drought in Northeast Brazil - past,
present, and future. Theor. Appl. Climatol. 129, 1189–1200. https://doi.org/
10.1007/s00704-016-1840-8.
Marengo, J.A., Alves, L.M., Ambrizzi, T., Young, A., Barreto, N.J.C., Ramos, A.M., 2020a.
Trends in extreme rainfall events and hydrogeometeorological disasters in the
metropolitan Area of São Paulo: a review. Ann. N. Y. Acad. Sci. 1472, 5–20. https://
doi.org/10.1111/nyas.14307.
Marengo, J.A., Cunha, A.P., Nobre, C.A., Neto, G.G.R., Magalhaes, A.R., Torres, R.R.,
Sampaio, G., Alexandre, F., Alves, L.M., Cuartas, L.A., Deusdará, K.R.L., Alvalá, R.C.
S., 2020b. Assessing drought in the drylands of northeast Brazil under regional
warming exceeding 4 C. Nat. Hazards 103, 2589–2611. https://doi.org/10.1007/
◦
s11069-020-04097-3.
Marengo, J.A., Camarinha, P.I., Alves, L.M., Diniz, F., Betts, R., 2021. Extreme rainfall
and hydro-geo-meteorological disaster risk in 1.5, 2.0, and 4.0 C global warming
◦
F. Jeferson de Medeiros et al.
scenarios: an analysis for Brazil. Frontiers in Climate 3, 610433. https://doi:10.338
9/fclim.2021.610433.
Martre, P., Wallach, D., Asseng, S., et al., 2015. Multimodel ensembles of wheat growth:
many models are better than one. Global Change Biol. 21, 911–925. https://do
i:10.1111/gcb.12768.
McPhillips, L.E., Chang, H., Chester, M., Depietri, Y., Friedman, E., Grimm, N.E.,
Kominoski, J.S., McPhearson, T., Méndez-Lázaro, P., Rosi, E.J., Shiva, J.S., 2018.
Defining extreme events: a cross-disciplinary review. Earth’s Future 6, 441–455. htt
ps://doi:10.1002/2017EF000686.
Medeiros, F.J., Oliveira, C.P., 2021. Dynamical aspects of the recent strong El Niño
events and its climate impacts in Northeast Brazil. Pure Appl. Geophys. 178,
2315–2332. https://doi.org/10.1007/s00024-021-02758-3.
Medeiros, F.J., Oliveira, C.P., 2022. Assessment of dry and heavy rainfall days and their
projected changes over Northeast Brazil in Coupled Model Intercomparison Project
Phase 6 models. Int. J. Climatol. https://doi.org/10.1002/joc.7759.
Medeiros, F.J., Oliveira, C.P., Torres, R.R., 2020a. Climatic aspects and vertical structure
circulation associated with the severe drought in Northeast Brazil (2012-2016).
Clim. Dynam. 55, 2327–2341. https://doi.org/10.1007/s00382-020-05385-1.
Medeiros, F.J., Oliveira, C.P., Santos e Silva, C.M., Araújo, J.M., 2020b. Numerical
simulation of the circulation and tropical teleconnection mechanisms of a severe
drought event (2012-2016) in Northeastern Brazil. Clim. Dynam. 54, 4043–4057.
https://doi:10.1007/s00382-020-05213-6.
Medeiros, F.J., Gomes, R.S., Coutinho, M.D.L., Lima, K.C., 2022. Meteorological drought
and water resources: historical and future perspective for Rio Grande do Norte state,
Northeast Brazil. Int. J. Climatol. https://doi.org/10.1002/joc.7624.
Meehl, G.A., Covey, C., Delworth, T., Mojib, L., McAvaney, B., Mitchell, J.F.B.,
Stouffer, R.J., Taylor, K.E., 2007. The WCRP CMIP3 multimodel dataset: a new era in
climate change research. Bull. Am. Meteorol. Soc. 88, 1383–1394. https://doi.org/
10.1175/BAMS-88-9-1383.
Montini, T.L., Jones, C., Carvalho, L.M.V., 2019. The South American low-level jet: a new
climatology, variability, and changes. J. Geophys. Res. Atmos. 124, 1200–1218.
https://doi.org/10.1029/2018JD029634.
Moriasi, D., Arnold, J., Van Liew, M.W., Bingner, R.L., Harmel, R.D., Veith, T.L., 2007.
Model evaluation guidelines for systematic quantifcation of accuracy in watershed
simulations. Trans. ASABE (Am. Soc. Agric. Biol. Eng.) 50, 885–900. https://doi.org/
10.13031/2013.23153.
Moss, R.H., Edmonds, J.A., Hibbard, K.A., Manning, M.R., Rose, S.K., van Vuuren, D.P.,
Carter, T.R., Emori, S., Kainuma, M., Kram, T., Meehl, G.A., Mitchell, J.F.B.,
Nakicenovic, N., Riahi, K., Smith, S.J., Stouffer, R.J., Thomson, A.M., Weyant, J.P.,
Willbanks, T.J., 2010. The next generation of scenarios for climate change research
and assessment. Nature 463, 747–756. https://doi.org/10.1038/nature08823.
Nakicenovic, N., Alcamo, J., Davis, G., De Vries, B., Fenhann, J., Gaffin, S., Gregory, K.,
Grubler, A., Jung, T.Y., Kram, T., La Rovere, E.L., Michaelis, L., Mori, S., Morita, T.,
Pepper, W., Pitcher, H., Price, L., Riahi, K., Roehrl, A., Rogner, H.H., Sankovski, A.,
Schlesinger, M., Shukla, P., Smith, S., Swart, R., Van Rooijen, S., Victor, N., Dadi, Z.,
2000. Special Report on Emissions Scenarios. Cambridge University Press,
Cambridge.
Neale, R.B., Richter, J.H., Jochum, M., 2008. The impact of convection on ENSO: from a
delayed oscillator to a series of events. J. Clim. 21, 5904–5924. https://doi.org/
10.1175/2008JCLI2244.1.
Nobre, C.A., Marengo, J.A., Seluchi, M.E., Cuartas, L.A., Alves, L.M., 2016. Some
characteristics and impacts of the drought and water crisis in southeastern Brazil
during 2014 and 2015. J. Water Resour. Protect. 8, 252–262. https://doi.org/
10.4236/jwarp.2016.82022.
Oliveira, P.T., Santos e Silva, C.M., Lima, K.C., 2017. Climatology and trend analysis of
extreme precipitation in subregions of Northeast Brazil. Theor. Appl. Climatol. 130,
77–90. https://doi.org/10.1007/s00704-016-1865-z.
Ortega, G., Arias, P.A., Villegas, J.C., Marquet, P.A., Nobre, P., 2021. Present-day and
future climate over Central and South America according to CMIP5/CMIP6 models.
Int. J. Climatol. 41, 6713–6735. https://doi.org/10.1002/joc.7221.
O’Neil, B.C., Tebaldi, C., van Vuuren, D.P., Eyring, V., Friedlingstein, P., Hurtt, G.,
Knutti, R., Kriegler, E., Lamarque, J.-F., Lowe, J., Meehl, G.A., Moss, R., Riahi, K.,
Sanderson, B.M., 2016. The scenario model intercomparison project (scenarioMIP)
for CMIP6. Geosci. Model Dev. (GMD) 9, 3461–3482. https://doi.org/10.5194/gmd-
9-3461-2016.
Pezzi, L.P., Quadro, M.F.L., Lorenzzetti, J.A., Miller, A.J., Rosa, E.B., Lima, L.N., Sutil, U.
A., 2022. The effect of Oceanic South Atlantic Convergence Zone episodes on
regional SST anomalies: the roles of heat fluxes and upper-ocean dynamics. Clim.
Dynam. https://doi.org/10.1007/s00382-022-06195-3.
Pfahl, S., O’Gorman, P.A., Fischer, E.M., 2017. Understanding the regional pattern of
projected future changes in extreme precipitation. Nat. Clim. Change 7, 423–427.
https://doi.org/10.1038/nclimate3287.
Rajão, R., Soares-Filho, B., Nunes, F., Börner, J., Machado, L., Assis, D., Oliveira, D.,
Pinto, L., Ribeiro, V., Rausch, L., Gibbs, H., Figueira, D., 2020. The rotten apples of
Brazil’s agribusiness. Science 369, 246–248. https://doi.org/10.1126/science.
aba6646.
Reboita, M.S., Ambrizzi, T., Silva, B.A., Pinheiro, R.F., Rocha, R.P., 2019. the South
atlantic subtropical anticyclone: present and future climate. Front. Earth Sci. 7, 8.
https://doi.org/10.3389/feart.2019.00008.
Reboita, M.S., Kuki, C.A.C., Marrafon, V.H., Souza, C.A., Ferreira, G.W.S., Teodoro, T.,
Lima, J.W.M., 2022. South America climate change revealed through climate indices
projected by GCMs and Eta-RCM ensembles. Clim. Dynam. 58, 459–485. https://doi.
org/10.1007/s00382-021-05918-2.
Regoto, P., Dereczynski, C., Chou, S.C., Bazzanela, A.C., 2021. Observed changes in air
temperature and precipitation extremes over Brazil. Int. J. Climatol. 41, 5125–5142.
https://doi.org/10.1002/joc.7119.
21
Weather and Climate Extremes 38 (2022) 100511
Reis, L.C., Santos e Silva, C.M., Bezerra, B.G., Mutti, P.R., Spyrides, M.H.C., Silva, P.E.,
2020. Analysis of climate extremes indices in the MATOPIBA region, Brazil. Pure
Appl. Geophys. 177, 4457–4478. https://doi.org/10.1007/s00024-020-02474-4.
Reis, J.S., Gonçalves, W.A., Mendes, D., 2021. Climatology of the dynamic and
thermodynamic features of upper tropospheric cyclonic vortices in Northeast Brazil.
Clim. Dynam. 57, 3413–3431. https://doi.org/10.1007/s00382-021-05873-y.
Ribeiro, M.S.M., Andrade, L.M.B., Spyrides, M.H., Lima, K.C., Silva, P.E., Batista, D.T.,
Lara, I.A.R., 2021. Environmental disasters in Northeast Brazil:
hydrometeorological, social, and sanitary factors. Weather, Climate, and Society 13,
541–554. https://doi.org/10.1175/WCAS-D-20-0132.1.
Rogelj, J., Meinshausen, M., Knutti, R., 2012. Global warming under old and new
scenarios using IPCC climate sensitivity range estimates. Nat. Clim. Change 2,
248–253. https://doi.org/10.1038/nclimate1385.
Santos, E.B., Lucio, P.S., Santos e Silva, C.M., 2015. Precipitation regionalization of the
Brazilian Amazon. Atmos. Sci. Lett. 16, 185–192. https://doi.org/10.1002/asl2.535.
Santos, E.B., Freitas, E.D., Rafee, S.B., Fujita, T., Rudke, A.P., Martins, L.D., Souza, R.A.
F., Martins, J.A., 2021. Spatio-temporal variability of wet and drought events in the
Paraná River basin-Brazil and its association with the El Niño-Southern oscillation
phenomenon. Int. J. Climatol. 41, 4879–4897. https://doi.org/10.1038/
nclimate1385.
Schumacher, R.S., Rasmussen, K.L., 2020. The formation, character and changing nature
of mesoscale convective systems. Nat. Rev. Earth Environ. 1, 300–314. https://doi.
org/10.1038/s43017-020-0057-7.
Sen, P.K., 1968. Estimates of the regression coefficient based on Kendall’s tau. J. Am.
Stat. Assoc. 63, 1379–1389. https://doi.org/10.2307/2285891.
Seneviratne, S., Hauser, M., 2020. Regional climate sensitivity of climate extremes in
CMIP6 versus CMIP5 multimodel ensembles. Earth’s Future 8, e2019EF001474.
https://doi.org/10.1029/2019EF001474.
Sheffield, J., Goteti, G., Wood, E., 2006. Development of a 50-year high-resolution global
dataset of meteorological forcings for land surface modeling. J. Clim. 19,
3088–3111. https://doi.org/10.1175/JCLI3 790.1.
Sillmann, J., Kharin, V.V., Zhang, X., Zwiers, F.W., Bronaugh, D., 2013a. Climate
extremes indices in the CMIP5 multimodel ensemble: Part 1. Model evaluation in the
present climate. J. Geophys. Res. Atmos. 118, 1716–1733. https://doi.org/10.1002/
jgrd.50203.
Sillmann, J., Kharin, V., Zwiers, W., Zhang, X., Bronaugh, D., 2013b. Climate extremes
indices in the CMIP5 multimodel ensemble: part 2. Future climate projections.
J. Geophys. Res. Atmos. 118, 2473–2493. https://doi.org/10.1002/jgrd.50188.
Silva, P.E., Santos e Silva, C.M., Spyrides, M.H.C., Andrade, L.M.B., 2019. Precipitation
and air temperature extremes in the Amazon and Northeast Brazil. Int. J. Climatol.
39, 579–595. https://doi:10.1002/joc.5829.
Silva, E.H.L., Silva, F.D.S., Silva Junior, R.S., Pinto, D.D.C., Costa, R.L., Gomes, H.B.,
Cabral Júnior, J.B., Freitas, I.G.F., Herdies, D.L., 2022. Perfomance assessment of
different precipitation databases (gridded analyses and reanalyses) for the new
Brazilian agricultural frontier: sealba. Water 14, 1473. https://doi.org/10.3390/
w14091473.
Skansi, M., Brunet, M., Sigró, J., Aguilar, E., Groening, J.A.A., Bentancur, O.J., Geier, Y.
R.C., Amaya, R.L.C., Jácome, H., Ramos, A.M., Rojas, C.O., Pasten, A.M., Mitro, S.S.,
Jiménez, C.V., Martínez, R., Alexander, L.V., Jones, P.D., 2013. Warming and
wetting signals emerging from analysis of changes in climate extreme indices over
South America. Global Planet. Change 100, 295–307. https://doi:10.1016/j.gloplach
a.2012.11.004.
Sperber, K.R., Annamalai, H., Kang, I.-S., Kitoh, A., Moise, A., Turner, A., Wang, B.,
Zhou, T., 2013. The Asian summer monsoon: an intercomparison of CMIP5 vs.
CMIP3 simulations of the late 20th century. Clim. Dynam. 41, 2711–2744. https://
doi:10.1007/s00382-012-1607-6.
Srivastava, A., Grotjahn, R., Ullrich, P.A., 2020. Evaluation of historical CMIP6 model
simulations of extreme precipitation over contiguous US regions. Weather Clim.
Extrem. 29, 100268 https://doi.org/10.1016/j.wace.2020.100268.
Taylor, K.E., Stouffer, R.J., Meehl, G.A., 2012. An overview of CMIP5 and the experiment
design. Bull. Am. Meteorol. Soc. 93, 485–498. https://doi.org/10.1175/BAMS-D-11-
00094.1.
Tebaldi, C., Knutti, R., 2007. The use of the multi-model ensemble in probabilistic
climate projections. Philos. Trans. R. Soc. A Math. Phys. Eng. Sci. 365, 2053–2075.
https://doi.org/10.1098/rsta.2007.2076.
Tebaldi, C., Hayhoe, K., Arblaster, J.M., Meehl, G.A., 2006. Going to the extremes. Clim.
Change 79, 185–211. https://doi.org/10.1007/s10584-006-9051-4.
Teixeira, M.S., Satyamurty, P., 2011. Trends in the frequency of intense precipitation
events in southern and southeastern Brazil during 1960-2004. J. Clim. 24,
1913–1921. https://doi.org/10.1175/2011JCLI3511.1.
Thibeault, J.M., Seth, A., 2014. Changing climate extremes in the Northeast United
States: observations and projections from CMIP5. Climatic Change 127, 273–287.
https://doi.org/10.1007/s10584-014-1257.
Torres, R.R., Marengo, J.A., 2013. Uncertainty assessments of climate change projections
over South America. Theor. Appl. Climatol. 112, 253–272. https://doi.org/10.1007/
s00704-012-0718-7.
Torres, R.R., Lapola, D.M., Marengo, J.A., Lombardo, M.A., 2012. Socioclimatic hotspots
in Brazil. Climatic Change 115, 597–609. https://doi.org/10.1007/s10584-012-
0461-1.
Trenberth, K.E., 1999. Conceptual framework for climate changes of extremes of the
hydrological cycle with climate change. Climatic Change 42, 327–339. https://doi.
org/10.1023/A100548890935.
Ukkola, A.N., Kauwe, M.G., Roderick, M.L., Abramowitz, G., Pitman, A.J., 2020. Robust
future change in meteorological drought in CMIP6 projections despite uncertainty in
precipitation. Geophys. Res. Lett. 47, e2020GL087820 https://doi.org/10.1029/
2020GL087820.
F. Jeferson de Medeiros et al.
Uvo, C.B., Repelli, C.A., Zebiak, S.E., Kushnir, Y., 1998. The relationship between
tropical Pacific and Atlantic SST and northeast Brazil monthly precipitation. J. Clim.
11, 551–562. https://doi.org/10.1175/1520-0442(1998)011<0551:TRBTPA>2.0.
CO;2.
Wehner, M., Gleckler, P., Lee, J., 2020. Characterization of long period return values of
extreme daily temperature and precipitation in the CMIP6 models: Part1, model
evaluation. Weather Clim. Extrem. 30, 100283 https://doi.org/10.1016/j.
wace.2020.100283.
Wild, M., 2020. The global energy balance as represented in CMIP6 climate models.
Clim. Dynam. 55, 553–577. https://doi.org/10.1007/s00382-020-05282-7.
Willmott, C., Robeson, S., Matsuura, K., Ficklin, D., 2015. Assessment of three
dimensionless measures of model performance. Environ. Model. Software 73,
167–174. https://doi.org/10.1016/j.envsoft.2015.08.012.
Wilson, A.B., Avila-Diaz, A., Oliveira, L., Zuluga, C., Mark, B., 2022. Climate extremes
and their impacts on agriculture across the eastern corn belt region of the. U.S.
Weather and Climate Extremes 37, 100467. https://doi.org/10.1016/j.
wace.2022.100467.
World Meteorological Organization (WMO), 2019. Atlas of Mortality and Economic
Losses from Weather, Climate and Water Extremes (1970-2019). WMO-No. 1267).
Wu, T., Lu, Y., Fang, Y., Xin, X., Li, L., Li, W., Jie, W., Zhang, J., Liu, Y., Zhang, L.,
Zhang, F., Zhang, Y., Wu, F., Li, J., Chu, M., Wang, Z., Shi, X., Liu, X., Wei, M.,
Huang, A., Zhang, Y., Liu, X., 2019. The beijing climate center climate system model
(BCC-csm): the main progress from CMIP5 to CMIP6. Geosci. Model Dev. (GMD) 12,
1573–1600. https://doi.org/10.5194/gmd-12-1573-2019.
Wyser, K., Noije, T., Yang, S., Hardenberg, J., O’Donnell, D., Döscher, R., 2020. On the
increased climate sensitivity in the EC-EARTH model from CMIP5 to CMIP6. Geosci.
Model Dev. (GMD) 13, 3465–3474. https://doi.org/10.5194/gmd-13-3465-2020.
Xavier, A.C., King, C.W., Scanlon, B.R., 2016. Daily gridded meteorological variables in
Brazil (1980-2013). Int. J. Climatol. 36, 2644–2659. https://doi.org/10.1002/
joc.4518.
Xu, K., Xu, B., Ju, J., Wu, C., Dai, H., Hu, B.X., 2019. Projection and uncertainty of
precipitation extremes in the CMIP5 multimodel ensembles over nine major basins in
China. Atmos. Res. 226, 122–137. https://doi.org/10.1016/j.atmosres.2019.04.018.
22
Weather and Climate Extremes 38 (2022) 100511
Yue, S., Pilon, P., Cavadias, G., 2002. Power of the Mann–Kendall and Spearman’s rho
tests for detecting monotonic trends in hydrological series. J. Hydrol. 259, 254–271.
https://doi.org/10.1016/S0022-1694(01)00594-7.
Zachariah, M., Vasconcelos Junior, F.C., Silva, T.L.V., Santos, E.P., Coelho, C.A.S.,
Alves, L.M., Martins, E.S.P.R., Köberle, A.C., Singh, R., Vahlberg, M., Marchezini, V.,
Heinrich, D., Thalheimer, L., Raju, E., Koren, G., Philip, S., Kew, S., Bonnet, R., Li, S.,
Yang, W., Sun, J., Vecchi, G., Otto, F.E.L., 2022. Climate Change Increased Heavy
Rainfall, Hitting Vulnerable Communities in Eastern Northeast Brazil.
Zanin, P.R., Satyamurty, P., 2020. Hydrological processes interconnecting the two largest
watersheds of South America from seasonal to intra-monthly time scales: a critical
review. Int. J. Climatol. 40, 3971–4005. https://doi.org/10.1002/joc.6443.
Zelinka, M.D., Myers, T.A., McCoy, D.T., Po-Chedley, S., Caldwell, P.M., Ceppi, P.,
Klein, S.A., Taylor, K.E., 2020. Causes of higher climate sensitivity in CMIP6 models.
Geophys. Res. Lett. 47, e2019GL085782 https://doi.org/10.1029/2019GL085782.
Zhang, X., Alexander, L., Hegerl, G., Jones, P., Tank, A.K., Peterson, T.C., Trewin, B.,
Zwiers, F.W., 2011. Indices for monitoring changes in extremes based on daily
temperature and precipitation data. Wiley Interdiscip. Rev. Clim. Chang. 2,
851–870. https://doi.org/10.1002/wcc.147.
Zhang, Y., You, Q., Chen, C., Ge, J., Adnan, M., 2018. Evaluation of downscaled CMIP5
coupled with VIC model for flash drought simulation in a humid Subtropical Basin,
China. J. Clim. 31, 1075–1090. https://doi.org/10.1175/JCLI-D-17-0378.1.
Zhu, H., Jiang, Z., Li, J., Li, W., Sun, C., Li, L., 2020. Does CMIP6 inspire more confidence
in simulating climate extremes over China? Adv. Atmos. Sci. 37, 1119–1132.
https://doi.org/10.1007/s00376-020-9289-1.
Zilli, M.T., Carvalho, L.M.V., Liebmann, B., Silva Dias, M.A., 2017. A comprehensive
analysis of trends in extreme precipitation over southeastern coast of Brazil. Int. J.
Climatol. 37, 2269–2279. https://doi.org/10.1002/joc.4840.
Zuluaga, C.F., Avila-Diaz, A., Justino, F.B., Martins, F.R., Ceron, W.L., 2022. The climate
change perspective of photovoltaic power potential in Brazil. Renew. Energy 193,
1019–1031. https://doi.org/10.1016/j.renene.2022.05.029.