Theoretical and Applied Climatology (2024) 155:7811–7826

https://doi.org/10.1007/s00704-024-05099-y

RESEARCH

Diurnal cycle of precipitation in Brazil

Cláudio Moisés Santos e Silva1,2 · Daniele Tôrres Rodrigues3 · Felipe Medeiros1 · Aléxia Monteiro Valentim1 ·
Paula Andressa Alves de Araújo1 · Joicy da Silva Pinto1 · Pedro Rodrigues Mutti1,2 · Keila Rêgo Mendes1 ·
Bergson Guedes Bezerra1,2 · Cristiano Prestrelo de Oliveira1,2 · Weber Andrade Gonçalves1,2

Received: 17 August 2023 / Accepted: 1 July 2024 / Published online: 11 July 2024
© The Author(s), under exclusive licence to Springer-Verlag GmbH Austria, part of Springer Nature 2024

Abstract
The diurnal cycle is an important mode of climatic variability associated with different aspects of micro, meso and large
scale meteorological phenomena. Thus, we performed a study of the space-time variability of the diurnal cycle of precipi-
tation with hourly sampling and covering all regions of Brazil. The dataset was collected during the period of 13-year,
from 1st January 2008 to 31th December 2020. We used data from 411 rain gauges installed in automatic weather sta-
tions. To evaluate regional aspects, we conducted a cluster analysis with different configurations (4, 5 and 6 groups). We
identified a considerable heterogeneity in the hour of maximum precipitation in Brazil and three main types of diurnal
cycle were observed: (i) maximum precipitation at mid- to late afternoon associated with strong local convection activity;
(ii) diurnal cycle with intense precipitation during nighttime at the Amazon basin, the coast of Northeast Brazil and the
Southern region; (iii) semidiurnal cycles with low precipitation rate at the Northeast Brazil.

1 Introduction are relevant scientific issues because it is present in many

environmental variables such as radiation, cloudiness, air,
The variability of the diurnal cycle is directly associated and sea surface temperature (Webster et al. 1996; Yan et al.
with surface forcing such as frictional drag, heat, mass, 2021), atmospheric lightning activity (Boccippio et al. 2002;
and energy fluxes. Thus, the diurnal cycle is the first mode Minobe et al. 2020), particulate matter and atmospheric pol-
of atmospheric variability affected by climatic changes, lution in urban and forested regions (Backman et al. 2012;
with possible implications on extreme events occurrence Wimmer et al. 2018).
(Meredith et al. 2019). It is difficult to study it through In continental areas, the diurnal cycle of precipitation
observational data collected in situ, since there are few is modulated by thermodynamic processes associated with
meteorological datasets with long term subdaily sampling in surface instability generated by solar heating during day
most regions of the globe, despite some incipient initiatives hours and by scale atmospheric dynamic, which is gener-
(e.g., Lewis et al. 2019). The systematic analysis of the diur- ally associated with low- and high-level jets over regions
nal cycle and the understanding of its physical mechanisms with strong convective activity (Tripoli and Cotton 1989;
Marengo et al. 2004; Giles et al. 2020). Additionally, low
level convergence, strong surface heating driving convec-
Cláudio Moisés Santos e Silva tive activity, orographic clouds formed by forcing lift or
[email protected] barometric tides can considerably influence on the diurnal
cycle of precipitation (Nesbitt and Zipser 2003; Janowiak
1
Programa de Pós-graduação em Ciências Climáticas, et al. 2005; Yang and Smith 2006; Yang et al. 2008; Watters
Departamento de Ciências Atmosféricas e Climáticas,
Universidade Federal do Rio Grande do Norte, Av. Senador et al. 2021).
Salgado Filho, 3000, Lagoa Nova, Natal 59078-970, Brazil The continental diurnal cycle of precipitation is typically
2
Departamento de Ciências Atmosféricas e Climáticas, characterized by a well pronounced maximum at mid- to late
Universidade Federal do Rio Grande do Norte, Lagoa Nova, afternoon, while in coastal and oceanic areas, the maximum
Brazil occurs during late-evening–early-morning hours. In this
3
Departamento de Estatística, Universidade Federal do Piauí, way, Yang and Smith (2006) performed a comprehensive
Av. Campus Universitário Ministro Petrônio Portella, Ininga, analysis, investigating the physical mechanisms controlling
Teresina 64049-550, Brasil

13
Content courtesy of Springer Nature, terms of use apply. Rights reserved.
7812 C. M. Santos e Silva et al.

the diurnal cycle of precipitation and classifying the diurnal and West areas, the annual precipitation ranges from 600
cycles as “oceanic or continental’’ and as “day- or night- to 900 mm yr− 1 (Rao et al. 2016; Rodrigues et al. 2020), in
type”. In addition, the authors indicated the regions with these regions Atlantic forest and Cerrado are found, respec-
prevalence of forced and natural convection in the tropical tively. Finally, in the Northwest (bordering AMZ), the pre-
regions. Notwithstanding, studies using different datasets cipitation reached 1800 mm yr− 1. These biomes found in
suggest the observation of a consistent subdaily cycle over NEB have experienced intense change associated with the
tropical and subtropical areas, including regions with differ- land use, in which the substitution of native vegetation by
ent climatic controls in the Amazon basin and other regions croplands is the main activity (Lapola et al. 2014; Campos
in South America (Yang et al. 2008; Santos and Silva 2013; et al. 2019; Reis et al. 2020). Concerning the diurnal cycle
Giles et al. 2020). precipitation in NEB, the maximum of precipitation can be
Due to the complexity of the physical processes modulat- observed at late afternoon into the continental and during
ing the diurnal cycle of precipitation, the simulations of this early morning in coastal areas (Kousky et al. 1980; Tenório
cycle have been used to evaluate the efficiency of climatic et al. 2010). These results obtained in coastal NEB were
models (da Rocha et al. 2009; Santos et al. 2012; Daleu et al. consistent with the oceanic mechanisms described by Yang
2020). In this way, some authors showed how the deficien- and Smith (2006). Further, there was strong influence of ter-
cies in cumulus parameterization or coarse resolution used restrial and sea breeze in the precipitation diurnal cycle in
in global and/or regional climate models can be crucial fac- coastal NEB (Souza and Oyama 2017).
tors representing the transition of shallow to deep convec- The Central-West of Brazil is dominated by Cerrado
tion (e.g., Bechtold et al. 2004). Furthermore, the accurate biome and the annual precipitation ranges from 544 to
representation of the timing and intensity of precipitation 2864 mm year− 1 (Spangler et al. 2017). In this region,
in models is directly associated with turbulence parameter- the precipitation is influenced by local convection and by
izations in the Atmospheric Boundary Layer (Kong et al. moisture transport from the Amazonia by events of South
2023), which directly influence the vertical distribution of Atlantic Convergence Zone (SACZ) during summer months
mass and energy in the atmosphere and, consequently, cloud (Carvalho et al. 2004), and due the incursion of polar air
formation (Lin et al. 2019). mass from mid latitudes, that reaches the central South
Furthermore, failure in the convective trigger function is America, favoring displacement of cold fronts during aus-
reported as a cause of difficulties in properly representing tral winter (Reboita et al. 2010; Rao et al. 2016). Despite
the diurnal cycle over continental areas (Itterly et al. 2016; the occurrence of cold fronts, the climate in central Brazil is
Pichelli et al. 2021; Cui et al. 2021). Moreover, studies ana- predominantly dry during austral winter and spring, with a
lyzing the difference between present and future climate typical migration of convection from SE to NW (Horel et al.
regarding the diurnal cycle of precipitation were performed, 1989), when the precipitation is associated with local con-
identifying changes in the precipitation intensity between vection, modulating the diurnal cycle of precipitation and
0300 and 1200 UTC around the world, including South the rainy season intensity (Prado et al. 2021).
America (Reboita et al. 2016; Watters et al. 2021), and asso- The Southeast region has the largest concentration of
ciating them with land use and land cover changes. people in urban areas. Studies indicate that the diurnal cycle
Brazil has continental dimensions and studies on the is associated with the growth of demographic and urban
diurnal cycle have been carried out using data obtained areas, implying microclimatic changes. For example, Lima
during experimental campaigns in different regions, such and Rueda (2018) verified an increase of diurnal tempera-
as the Amazon basin (AMZ) (Marengo et al. 2004; San- ture range in São Paulo city. Many meteorological systems
tos et al. 2009; Biscaro et al. 2021). In this perspective, the influence the precipitation distribution in the Southeast
analysis showed the influence of land use, which is associ- region. In this way, Reboita et al. (2010) described the
ated to transition of natural biomes to pasture, can modi- relative role of cold fronts, high level cyclonic vortex and
fied the hydrologic cycle, thermodynamic structure of low the SACZ over the rainfall observed, showing that these
troposphere and, consequently, in the diurnal variability of meteorological systems can occur all year long. However,
cloudiness and the precipitation regimes in AMZ (Wang et the rainy season is observed during the austral summer
al. 2000, 2009; Silva Dias et al. 2002; Durieux et al. 2003; months (Rao et al. 2016). The diurnal cycle in Southeast is
Khanna et al. 2018). well pronounced, with maximum occurring in mid- to late
The Northeast of Brazil (NEB) there is a high temporal afternoon, promoting strong atmospheric lightning activity
and spatial variability of precipitation. In the Semiarid area, (Mattos et al. 2017).
where Caatinga is the main biome, the annual precipitation In Brazil, the Low-Level Jets (LLJ) East of the Andes
ranges from 300 to 600 mm yr− 1 (Oliveira et al. 2017; Cam- is an important mechanism modulating the diurnal cycle
pos et al. 2019; Mutti et al. 2019). In the Eastern coastal of precipitation in many regions of Brazil (Marengo et al.

13
Content courtesy of Springer Nature, terms of use apply. Rights reserved.
Diurnal cycle of precipitation in Brazil 7813

2004; Oliveira et al. 2018; Giles et al. 2020). The LLJ has continental areas of Brazil. In this sense, our aim is to
influenced the time-space distribution of precipitation over characterize the time-space aspects of diurnal cycle of pre-
continental South America (Marengo et al. 2004; Montini et cipitation in Brazil during a continuous period of 13 years.
al. 2019). For example, in South region of Brazil and South Additionally, we demonstrate the regional aspects of the
America, the LLJ can be associated with the Subtropical Jet diurnal cycle using multivariate analysis.
Stream resulting in Mesoscale Convective Complex (MCC)
mainly at nighttime during austral summer (Marengo et al.
2004; Durkee et al. 2009). 2 Materials and methods
Recently the diurnal cycle of precipitation throughout
Brazil was analyzed via remote sensing retrievals limited 2.1 Database
to a 5-year (2014–2018) period of analysis (Afonso et al.
2020; Siqueira et al. 2021). Additionally, they did not ana- Brazil territory is the fifth largest country covering 8,547,403
lyze in a systematic way the whole Brazilian territory, but km² of which approximately 50% is composed of the Ama-
regions with well-defined diurnal cycles of precipitation. zonia region. The database used was obtained from the
In another perspective, using reanalysis dataset, Giles et al. “Meteorological Dataset for Teaching and Research’’ proj-
(2020) studied the rainfall diurnal cycle in South America ect of the INMET (www.inmet.gov.br). The dataset is com-
but without a specific look for the specific characteristics of posed of hourly precipitation collected by 411 automatic
Brazilian territory. weather stations covering the January 1, 2008 to December
From January 2008 onwards, the National Institute of 31, 2020 period. A comprehensive dataset obtained by 411
Meteorology (Instituto Nacional de Meteorologia - INMET) rain gauges installed in automatic weather stations distrib-
monitors different meteorological variables with sampling uted over the entire Brazilian territory was used (Fig. 1).
of one hour through automatic weather stations over all The climatic types according to Köppen’s classification

Fig. 1 Geographical location of Brazil and spatial distribution of auto- of Brazil are indicated: North, Northeast, Central-West, Southeast and
matic weather stations used in this study and the Köppen Climate clas- South, respectively
sification performed by (Alvarez et al., 2014). Additionally the regions

13
Content courtesy of Springer Nature, terms of use apply. Rights reserved.
7814 C. M. Santos e Silva et al.

(Alvarez et al. 2013) are also depicted in Fig. 1, where the Cluster analysis is a multivariate statistical method aimed
majority of Brazil’s continental territory is dominated by at partitioning a sample or population into groups such that
type A climates (tropical climates, including savannas) in elements within a group are similar and elements across
the Amazon, Central-West, and Northeast regions; while the groups are dissimilar with respect to the involved vari-
Southeast and South portions are characterized by type C ables, thereby maximizing within-group homogeneity and
(humid subtropical climates). Additionally, the majority of between-group heterogeneity (Mingoti 2005).
the Northeast exhibits a hot Sem-iarid climate (Bsh). The initial step in clustering involves computing a simi-
The data was collected by a sampling of 5 min later con- larity or dissimilarity measure; this study employs the
verted to 1-hour accumulated rainfall. Rainfall is described Euclidean distance, as previously used by Rodrigues et al.
in terms of precipitation rate (mm/hour). The full dataset has (2020), to generate a distance matrix for daily precipita-
1.1% of missing data. Although it is possible to fill in miss- tion values across municipalities in Northeast Brazil. The
ing data through multiple imputation methods (Rodrigues et Euclidean distance between two elements Xl and Xk, with
al. 2023), we chose to work with the original data, as even l ≠ k, is defined by Eq. 1.
multivariate methods tend to perform poorly in attempting
 12
to impute zero-inflated data, which is the case with hourly
 p
(Xil − Xik )  (1)
   12  2
d (Xl , Xk ) = (Xl − Xk ) (Xl − Xk ) =
precipitation time series (Ancelet et al. 2010). i=1

2.2 Clustering method In which Xl and Xk represents the two elements being com-
pared, and 𝑝 denotes the number of variables; i.e., the two
Due to the large extension of Brazil the diurnal cycle of sample elements are compared across each variable in the
precipitation was analyzed through the definition of homo- observation vector.
geneous groups. In this way, we used clustering analysis, The second stage of the clustering process involves tech-
which is a multivariate statistical technique aiming to sepa- niques for cluster construction (Mingoti 2005). We used
rate unequal elements of a sample or population. Once sepa- Ward’s hierarchical approach, proposed by Ward (1963)
rated, the method aggregates elements with homogeneous and used by various authors (Rodrigues et al. 2020; Silva
characteristics, maximizing homogeneity within the groups, et al. 2023). In this method, each element is initially con-
while maximizing heterogeneity among the groups (Mingoti sidered a single group, and at each step, the sum of squares
2005). The first stage of the clustering process is the estima- within each group is calculated. This sum is the square of
tion of a measure of similarity (or dissimilarity) among the the Euclidean distance from each element in the group to the
data and, we used the Euclidean distance which according group’s mean vector, represented by Eq. 2.
to Mimmack et al. (2001). This metric has been used for the
ni 
regionalization of climate in many climatic studies in Brazil    
(Oliveira et al. 2017; Mutti et al. 2019).
SSi = Xij − X
−
Xij − X
−  (2)
i. i.
j=1
The second step is to define the linkage method, which
can be classified into agglomerative and divisive (Ward
1963). Five clustering heuristics options are available: In which ni is the number of elements in group 𝐺𝑖 at step 𝑘
single, centroid, complete, average and Ward’s. We used of the clustering process, 𝑋𝑖j is the observation vector of the
the hierarchical agglomerative Ward method which is j-th element in the i-th group, 𝑋i is the centroid of group Gi
based on the minimum variance inside a cluster (Hervada- and SSi represents the corresponding sum of squares for Gi.
Sala and Jarauta-Bragulat 2004). The clustering method At step 𝑘, the total within-group sum of squares is defined
was performed for four, five and six groups. The analysis by Eq. 3.
was performed in different phases. Firstly, we showed the
gk
diurnal cycle for annual average, considering the three 
SSR = SSi  (3)
clustering configurations (4, 5 and 6 groups). Secondly,
i=1
the clustering analysis was presented in quarters repre-
senting each season: DJF (December-January-February), in which gk is the number of existing groups at step k.
MAM (March-April-May), JJA (June-July-August), SON The distance between clusters 𝐺l and Gi is defined by
(September-October-November). Eq. 4, representing the sum of squares between clusters Gl
To conduct clustering analysis, we used the hourly pre- and Gi. At each step of the clustering algorithm, the two
cipitation rates from all 411 weather stations during the groups that minimize this distance are combined:
period from January 1, 2008, to December 31, 2020 as vari-
ables in the Ward method.

13
Content courtesy of Springer Nature, terms of use apply. Rights reserved.
Diurnal cycle of precipitation in Brazil 7815

 
nl ni  3 Results and discussion
  
d (Gl , Gi ) = X
−
− X
−
X−
− X
−  (4)
nl + ni l. i. l. i.

3.1 Annual average of precipitation diurnal cycle

The aim is to determine a number of groups that naturally
partition the elements under analysis (Mingoti 2005). In this Figure 2 shows the sum of squared distances for k in the
sense, the cophenetic correlation coefficient (CCC) helps range 1–11. If the plot looks like an arm, then the elbow on
assess the quality of these partitions by comparing the origi- the arm is ideal k. As k increases, the sum of the squared
nal distance matrix with the cophenetic matrix, reflecting distance tends to zero. When we set k to its maximum value
how well the dendrogram preserves original data distances n (where n is the number of samples) each sample will form
(Sokal and Rohlf 1962). A high CCC suggests an accurate its own cluster, which means that the sum of the squared
cluster configuration, while a low CCC indicates insufficient distances is equal to zero. In the figure above the elbow is at
modeling of data relationships. Furthermore, the elbow plot k = 7 indicating the optimal k for this dataset is 7.
helps identify the optimal number of clusters by showing The annual average of diurnal cycles of precipitation
the change in within-group sum of squares relative to the for three cluster configurations (4, 5, and 6 clusters) were
number of groups; a significant change suggests the ideal shown in Fig. 3. Four main groups of diurnal variation in
cluster count (Ketchen and Shook 1996; Onumanyi et al. precipitation were identified, distributed across the conti-
2020). This approach has been employed by researchers like nental territory of Brazil (Fig. 2a), with their average cycles
Amorim et al. (2020) and Silva et al. (2023) in environmen- depicted in Fig. 2b. The main characteristics of these groups
tal variable clustering. are as follows: Group 1 - maximum precipitation in the late
afternoon, extending into the night, with no precipitation
in the early morning; Group 2 - maximum precipitation in
the late afternoon, extending into the night, with precipita-
tion present in the morning; Group 3 - rain concentrated in

Fig. 2 Sum of squared distances for k values 1–11

13
Content courtesy of Springer Nature, terms of use apply. Rights reserved.
7816 C. M. Santos e Silva et al.

Fig. 3 Diurnal cycle of precipitation for different configuration of groups in Brazilian territory

the late night and early morning; Group 4 - bimodal pattern 6 - very intense peak precipitation recorded approximately
with rains concentrated in the late afternoon and early morn- an hour before the main peak of Group 1.
ing. In addition to the four main groups, two other groups Group 1 is most frequently observed in the Southern
(which are variations of Group 1) were identified: Group and eventually in the Southeastern and Central-Western
5 - maximum precipitation in the late afternoon and early of Brazil. Group 2 is predominantly located in Central-
night, with higher intensities compared to Group 1; Group Western region. The main difference between groups 1 and
2 is the intensity of maximum precipitation, which is most

13
Content courtesy of Springer Nature, terms of use apply. Rights reserved.
Diurnal cycle of precipitation in Brazil 7817

pronounced in the Southern region. The group 3 is observed Plata basin (Salio et al. 2007; Moraes et al. 2020) during the
in the coast of NEB, with a maximum of 0.30 mm/h at 0900 night times period (Giles et al. 2020).
UTC. The group 4 is characterized by a bimodal behavior The group 3 was observed in two regions with different
with two maximums at 0700 UTC and 1900 UTC. Finally, climate regimes. In AMZ, the nighttime rainfall is explained
in AMZ, despite the predominance of type 1 (intense precip- by the Squall Lines (from the Coastal AMZ) that propagates
itation at mid- to late afternoon), we can observe a mixing into the continent, consistent with previous studies (Cohen
of diurnal cycles, with precipitation observed during both, et al. 1995; Alcântara et al. 2011; Santos and Silva 2013),
night and early morning periods. including these meteorological systems are associated with
Furthermore, the main peak of precipitation in the late the extreme rainfall event during the nighttime period (San-
afternoon, modulated by the local convection mechanism, tos et al. 2016). From the perspective of large-scale atmo-
is associated with local convection. The hourly rates are less spheric dynamics, the diabatic heating due the condensation
than 0.35 mm/h when considering 4 or 5 groups. However, of water vapor in AMZ modulates the divergence at high
with 6 groups, a maximum of 0.42 mm/h was observed levels over the region as suggested by numerical and obser-
(group 6) at 1900 UTC, corresponding to the coastal and vational analysis (Silva Dias et al. 1987). The secondary
inland regions in the AMZ. maximum of nighttime precipitation was also reported in
Group 5, which presents a maximum at 2100UT, is previous studies with rainfall information from multiple
located basically in two regions: (i) parallel to the coastal sources (rain gauges, weather radar, and satellites) con-
North between 200 and 400 km into the continent; (ii) con- ducted during the rainy season in the Southwest of the basin
centrated in both, Southeast and coastal South of Brazil. The (Santos et al. 2009; Biscaro et al. 2021).
intensity of precipitation in group 5 covers the nighttime In the Coastal NEB, the group 3 is typical of regions influ-
hours, decreasing from 2200 UTC onwards. The group 6 enced by proximity with oceans as described in Yang and
exhibits semidiurnal characteristics observed in AMZ, with Smith (2006) and the rainfall hours and intensity is modu-
a well defined maximum of 0.42 mm/h observed at 1900 lated by sea breeze mechanism as suggested by Gray and
UTC, while the secondary of 0.20 mm/h was observed at Jacobson (1977) and Ruppert and Kloche (2019). Finally,
0900 UTC. the diurnal cycle of semiarid region (group 4) is character-
On a preliminary analysis, the predominance of diurnal ized by low precipitation intensity and by a semidiurnal pat-
cycles associated with strong convective activity at mid- to tern, being the maximum at morning hours associated with
late afternoon in Brazil is because most of the territory is in oceanic mechanisms (similar to the group 3) and the second-
tropical/continental areas. Moreover, the presence of large ary maximum influenced by local convection. These results
rivers and forested areas supply moisture and heat to the are consistent with classical analysis of Kousky (1980), by
atmosphere through evapotranspiration phenomena, influ- studies conducted using weather radar information (Tenório
encing the formation and development of deep convection as et al. 2010), as well as simulations with numerical models
suggested by the classical mechanisms described by Ogura (Souza and Oyama 2017).
and Takahashi (1971). In this way, the convective activity
exhibited the strongest intensity at mid- to late afternoon as 3.2 Seasonal aspects of precipitation diurnal cycle
confirmed by different analysis performed previously from
the rain gauges and radar datasets in AMZ (Santos et al. The diurnal cycle of quarter DJF (Fig. 4) maintains some
2009; Biscaro et al. 2021). The maintenance of convection aspects of annual average precipitation; however, the group
during daytime hours is also associated with the thermal tur- 3 does not present the maximum precipitation at 0900 UTC.
bulence, inducing instability in low tropospheric levels and In the 4-group configuration, groups 1 and 2 exhibit a maxi-
intensifies the local convection following the mechanisms mum of 0.45 mm/h and 0.60 mm/h in the late afternoon,
described by Wallace (1975). but a secondary maximum between 06h00 and 09h00 in
In the South region, group 2 is the diurnal cycle preva- group 2 was observed. Finally, two bimodal groups (3 and
lent and exhibits two maximum rainfall rates. The first at 4) are identified, the group 4 are concentrated in the semi-
mid- to late afternoon can be explained by the local convec- arid region of Brazil and the group 3 is spread across Brazil,
tion mechanism early described. The secondary maximum, standing out in the southern region and the semiarid and
during late-evening–early-morning hours, is influenced by Central-West transition areas. In the 5-groups configuration,
large scale mechanisms, notably the association between the group 5 shows a well pronounced maximum of 0.50 mm/h
East Andes LLJ (Marengo et al. 2004; Montini et al. 2019) but is delayed by two hours relative to group 1.
with the High-Level Subtropical Jet (Salio et al. 2007), Group 5 is basically concentrated in the Southeast region,
inducing baroclinicity and influencing the formation and when the low-level convergence modulates the SACZ
maintenance of Mesoscale Convective Complexes in La occurrence. In another way, group 1 is most frequent in the

13
Content courtesy of Springer Nature, terms of use apply. Rights reserved.
7818 C. M. Santos e Silva et al.

Fig. 4 Similar to Fig. 3 but for quarter DJF

South region of Brazil. The 6-groups configuration creates to AMZ region (Carvalho et al. 2004; Raia and Cavalcanti
a new group of nocturnal rainfall distributed in Central of 2008; Marengo et al. 2012). The SACZ intensity modulates
Brazil and AMZ with a maximum of 0.50 mm/h, localized the diurnal rainfall in the groups 1 and 2, with a maximum
across SACZ areas. The meteorological aspects during the of precipitation at late-afternoon influencing the convec-
quarter DJF in Brazil are consistent with the presence of tive activity of both groups (1 and 2), which are found from
South America Monsoon System (SAMS), in which the Southeast, Central of Brazil and AMZ.
SACZ exhibit the SE-NW orientation from the Southeast

13
Content courtesy of Springer Nature, terms of use apply. Rights reserved.
Diurnal cycle of precipitation in Brazil 7819

In DJF the group 6 exhibits a nocturnal precipitation cycle of maximum at late afternoon. In another way, in the
located in inland areas of AMZ, this maximum can be Central-East, Southeast and South region, we identified diur-
explained due the SL propagation, which are observed dur- nal cycles with less rainfall rates as suggested, for example,
ing all year penetrating into the continent (Cohen et al. 1995; the study of Rao et al. (2016). Additionally, the diurnal cycle
Alcântara et al. 2011). We have identified an expansion of with nighttime rainfall in the coastal NEB (group 4) indi-
area covering by group 4 toward the coastal region; how- cates the activity of Easterly Wave Disturbances according
ever, the bimodal aspect of diurnal cycle was maintained, to Gomes et al. (2019), which are influenced by the increase
corroborating with previous studies over the region (Kousky in sea surface temperature and, therefore, can be explained
1980; Souza and Oyama 2017). This bimodal feature is due through the mechanism proposed by Sui et al. (1997). How-
to the proximity to the oceanic regions as reported for other ever, it is not possible to directly attribute this precipitation
tropical coastal regions of the globe (Yang et al. 2008). The to the occurrence of EWD, as it would require an analysis of
group 3 exhibited maximum in the late afternoon but with case compositions of such systems, simultaneously allow-
less intensity compared to groups 1, 2 and 5. ing us to analyze the partitioning of precipitation in terms
Despite the similarity among the diurnal cycles, the of convective and stratiform clouds. This methodology can
meteorological features of the late-afternoon maximum are be conducted using satellite products or through numeri-
different. In the Borborema Plateau, located in Semiarid of cal models of the climate system; however, it is beyond the
Brazil, the moisture air from the South Atlantic tropical with scope of the present research.The nighttime precipitation of
direction East-Southeast is raised when passing through by group 4 has been reported as a cause of natural disasters
topography influencing the formation of cumulus clouds as and consequent economic and health implication in some
suggested by the classical mechanism described by Tripoli subregions (mostly coastal) of NEB (Rodrigues et al. 2020,
and Cotton (1989). In another way, the NEB is influenced 2021).
by large scale meteorological systems, which promotes pre- In the quarter JJA (Fig. 6) the regions South (group 4)
cipitation (at any time) during the quarter DJF, specifically and Southeast (group 2) stand out for presenting diurnal
the Upper Tropospheric Cyclonic Vortices are observed in cycle with low amplitude along the 24 h and intensity vary-
this period (e.g., Dos Reis et al. 2021). In the inland areas ing between 0.2 mm/hour e 0.1 mm/hour, respectively. At
of AMZ and in the transition of the semiarid to Southeast same time, all Central East and the most part of Semiarid
region the local convection mechanism (Wallace 1975) is shows the lowest precipitation rates, being approximately
dominant. In the South, which is a subtropical region of zero throughout the day, thus during JJA are observed the
Brazil, the incoming solar radiation during the summer lowest precipitation accumulation in the Brazilian territory
months causes strong heating and consequently instability (Rao et al. 2016), although with more frequent occurrences
in atmospheric boundary layer, this mechanism favors the of tropical cyclones and frontal systems reaching the South
formation of precipitation at late afternoon as described by region (Evans and Braun 2012; Escobar et al. 2019), what
Ogura and Takahashi (1971) and Wallace (1975). modulates the diurnal cycle of group 4.
The quarter MAM (Fig. 5) is characterized by a consider- The groups 3, 5 and 6 show diurnal cycle with maximum
able homogeneity in the diurnal cycle of precipitation. Group pronounced. Group 3 is persistent and located in the coastal
1 was the most frequently observed, in which the maximum areas of AMZ with a maximum rate of 0.42 mm/h at 1800
precipitation occurs at 1900 UTC with a rate of 0.25 mm/h, UTC and associated with squall line formation and propa-
being predominant in most of central Brazil, inland of NEB, gation as previously described. The South of AMZ is influ-
South and Southeast. In the coast of NEB, the diurnal cycle enced by dry and cold air mass (also called of friagens),
is typically characterized by nocturnal occurrence of rain- which are eventually associated with the penetration of
fall (0600UTC). In AMZ three cycles were observed. The front cold from the mid latitudes of South Hemisphere, sta-
first, with well pronounced precipitation at late- afternoon bilizing the convective boundary layer and decreasing the
(group 3) and maximum intensity of 0.85 mm/h. The sec- convective precipitation in this region (Fisch et al. 2004;
ond, concentrated in areas away from the coastal zone, in Camarinha-Neto et al. 2021) as suggested by the Ogura
which nocturnal precipitation is observed. The third, with and Takahashi (1971) and Wallace (1975) mechanisms. It
intensity of 0.35 mm/h at 1800 UTC, covering the inland is worth noting that, on average, 3 to 5 friagens events are
areas of AMZ. Finally, the configuration with 6 clusters is observed per year in the South of the AMZ. Therefore, the
characterized by the differentiation between groups 6 and 3, physical mechanisms described by these classic authors may
which have similar intensities, but with a two-hour delay. play a secondary role in modulating the diurnal cycle in this
Concerning the climatological aspects, the quarter MAM region. On the other hand, the region is also influenced by
shows high precipitation rates in Eastern AMZ and North- the formation of the SACZ, which modulates the intensity
ern NEB (de Souza et al. 2005), consistent with the diurnal

13
Content courtesy of Springer Nature, terms of use apply. Rights reserved.
7820 C. M. Santos e Silva et al.

Fig. 5 Similar to Fig. 3 but for quarter MAM

of precipitation and cloud cover in the area, as suggested by when the sea breeze is established as described by Souza
previous studies (Carvalho et al. 2004; Pezzi et al. 2023). and Oyama (2017). Group 6 is a variant of group 3 but with
Group 5 covers the coastal area of NEB. Specifically a maximum 2-hour delay, located in coastal North and, in
the East region (between 12oS e 5oS) is affected by East- some cases, inside the AMZ. The maximum intensity is
erly Wave Disturbances more frequently during this quar- 0.4 mm/h and the precipitation occurs in most cases in the
ter (Gomes et al. 2019). In this region, the diurnal cycle is late-afternoon.
characterized by occurrence of rainfall during all nighttime,

13
Content courtesy of Springer Nature, terms of use apply. Rights reserved.
Diurnal cycle of precipitation in Brazil 7821

Fig. 6 Similar to Fig. 3 but for quarter JJA

The diurnal cycle observed during the quarter SON to the rest of the country. Southeast, Central-West and AMZ
(Fig. 7) is the most heterogeneous, despite the regions South are characterized by diurnal cycles well defined at late-
and North of NEB can be highlighted. In the South, a diurnal afternoon. Group 1, concentrated in the Southeast region,
cycle without preferred time was observed; however, there the maximum precipitation occurs between 1800 and 2200
are two smooth maximums of 0.25 mm/h at 1000 and 1500 UTC with rates of 0.27 mm/h, while group 2 (distributed
UTC. In parts of the AMZ and NEB (excluding the coastal heterogeneously) shows a maximum precipitation rate of
East) the precipitation observed was smaller by comparison 0.15 mm/h at 1900 UTC. In the configuration with 5 and 6

13
Content courtesy of Springer Nature, terms of use apply. Rights reserved.
7822 C. M. Santos e Silva et al.

Fig. 7 Similar to Fig. 3 but for quarter SON

groups, we observe the formation of group 5, which has a 4 Conclusions

maximum of 0.4 mm at 1900 UTC, concentrated on North-
western AMZ and the group 6 is a variation, in terms of We showed a comprehensive analysis of the diurnal cycle
intensity, of the groups 1 and 2. of precipitation in continental territory of Brazil from an
in-situ database collected with a dense network of weather
automatic stations during the period from 1st January
2008 to 31th December 2020. After applying multivariate

13
Content courtesy of Springer Nature, terms of use apply. Rights reserved.
Diurnal cycle of precipitation in Brazil 7823

Acknowledgements To the National Council for Scientific and Tech-

analysis to determine different groups of diurnal variability, nological Development (CNPq) for the productivity grant conceded to
characterizing the discussing regionals and general aspects the first author (grant 310781/2020-5).
of diurnal cycle in Brazil, as well as the mechanisms modu-
late them. Author contributions Conceptualization: C.M.S.S., D.T.R., B.G.B.
Considering the annual average, four groups was identi- Methodology: C.M.S.S, D.T.R., F.M.; K.R.M. Software: D.T.R, F.M.
Formal analysis: C.M.S.S., B.G.B., P.R.M., C.P.O., W.A.G., J.S.P.,
fied: (i) diurnal cycle with maximum of precipitation intense A.M.V., D.T.R., P.A.A.A. Resources: C.M.S.S., B.G.B., P.R.M. Data
at late afternoon (groups 1 and 2 of the Fig. 2); (ii) diurnal Curation: D.R.T. Writing - Original Draft: C.M.S.S., D.R.T., F.M.,
cycle with intense precipitation during the nighttime period K.R.M., B.G.B. Writing - Review & Editing: C.M.S.S. Visualization:
(group 3 of the Fig. 2); (iii) group 4 of the Fig. 2, character- D.R.T., C.M.S.S., F.M. Supervision: C.M.S.S. Project administration:
C.M.S.S.
ized by a semidiurnal behavior and low precipitation rates.
It is possible to identify other diurnal cycles (maximum of 6,
Declarations
as shown in the Fig. 2), but the difference is concerning the
intensity of maximum and/or a delay of one or two hours by Competing interests I state herewith that the material of this article
comparing to the group 1. is not being presently considered for publication in any other journal,
In terms of spatial distribution, we can observe that diur- that its publication is approved by all authors and by the responsible
authorities where the work was carried out, and that, if accepted, it will
nal cycle with maximum precipitation at nighttime were
not be published elsewhere, including electronically, in the same form,
concentrated in East coastal of NEB and in the inland areas in English or in any other language, without the written consent of the
of AMZ; in the coastland of NEB the sea breeze mechanism copyright-holder.
has an important meteorological mesoscale influence on the
time of precipitation, while in the inland AMZ the propa-
gation of the squall lines is associated with the nighttime References
occurrence of precipitation in that region.
The seasonality has influence in the spatial distribution Afonso JMS, Vila DA, Gan MA, Quispe DP, Barreto NdJdC JH, Palha-
rini RSA (2020) Precipitation Diurnal Cycle Assessment of Satel-
of clusters. During the quarters DJF and SON was observed
lite-Based Estimates over Brazil. Remote Sensing, 12(14):2339.
a greater heterogeneity of diurnal cycles compared to the https://doi.org/10.3390/rs12142339
quarter MAM and JJA. Specifically, during the quarter Alcântara CR, Silva Dias MAF, Souza EP, Cohen JCP (2011) Verifica-
MAM, the most part of continental areas of Brazil show a tion of the role of the low level jets in Amazon squall lines. Atmos
Res 110:36–44. https://doi.org/10.1016/j.atmosres.2010.12.023
diurnal cycle with low precipitation rates and maximum at
Alvares CA, Stape JL, Sentelhas PC, Gonçalves JDM, Sparovek G
late afternoon. In JJA, only the coastal areas of NEB and (2013) Köppen’s climate classification map for Brazil. Meteorol
inland areas of AMZ exhibited diurnal cycles with well pro- Z 22(6):711–728. https://doi.org/10.1127/0941-2948/2013/0507
nounced maximum in nighttime and late afternoon hours, Amorim ACB, Lucio PS, Santos e Silva CM, Rodrigues DT, Brito
FAD (2020) Regionalização E análise da tendência da precipi-
respectively.
tação do Rio Grande do Norte associados a padrões de TSM.
Local factors have strong impacts on precipitation Revista Brasileira De Meteorologia 35(2):269–280. https://doi.
dynamics over such a heterogeneous area as the continen- org/10.1590/0102-7786352024
tal region of Brazil. However, in terms of the diurnal cycle, Ancelet S, Etienne MP, Benoît H, Parent E (2010) Modelling spatial
zero-inflated continuous data with an exponentially compound
what is basically observed, even in global studies (Yang and
Poisson process. Environ Ecol Stat 17(3):347–376. https://doi.
Smith 2006), is a diversity of mechanisms that control the org/10.1007/s10651-009-0111-6
diurnal cycle, but not a diversity of “modes” of variability. Backman J, Rizzo LV, Hakala J, Nieminen T, Manninen HE, Morais
In other words, in practice, rainfall either occurs in the late F, Kulmala M (2012) On the diurnal cycle of urban aero-
sols, black carbon and the occurrence of new particle forma-
afternoon, late at night, or early in the morning (or may be a
tion events in springtime São Paulo, Brazil. Atmos Chem Phys
mixture of these modes). On the other hand, rainfall intensi- 12(23):11733–11751
ties vary greatly, and from this perspective, certainly a more Bechtold P, Chaboureau JP, Beljaars A, Betts AK, Köhler M, Miller
thorough analysis can be conducted in future opportunities M, Redelsperger JL (2004) The simulation of the diurnal cycle
of convective precipitation over land in a global model. Q J
to discuss what causes changes in cycle intensities in light
Royal Meteorological Society: J Atmospheric Sci Appl Meteo-
of anthropogenic influence. rol Phys Oceanogr 130(604):3119–3137. https://doi.org/10.1256/
Finally, we emphasize that the 13-year period is not long qj.03.103
enough to represent the climatology of the diurnal precipita- Biscaro TS, Machado LA, Giangrande SE, Jensen MP (2021) What
drives daily precipitation over the central Amazon? Differences
tion cycle over Brazil, as from the classical perspective, we
observed between wet and dry seasons. Atmos Chem Phys
would require a minimum of 30 years. Therefore, this study 21(9):6735–6754. https://doi.org/10.5194/acp-21-6735-2021
should not be considered from a climatological perspective. Boccippio DJ, Koshak WJ, Blakeslee RJ (2002) Perfor-
Thus, updates of these analyses will be necessary in future mance assessment of the optical transient detector
and lightning imaging sensor. Part I: predicted diurnal
opportunities.

13
Content courtesy of Springer Nature, terms of use apply. Rights reserved.
7824 C. M. Santos e Silva et al.

variability. J Atmos Ocean Technol 19(9):1318–1332. https:// Gomes HB, Ambrizzi T, Pontes da Silva BF, Hodges K, Silva Dias
doi.org/10.1175/1520-0426(2002)019<1318:PAOTOT>2. PL, Herdies DL, Gomes HB (2019) Climatology of easterly wave
0.CO;2 disturbances over the tropical South Atlantic. Clim Dyn 53:1393–
Camarinha-Neto GF, Cohen JC, Dias-Júnior CQ, Sörgel M, Cattanio 1411. https://doi.org/10.1007/s00382-019-04667-7
JH, Araújo A, Artaxo P (2021) The friagem event in the central Gray WM, Jacobson RW Jr (1977) Diurnal variation of deep cumu-
Amazon and its influence on micrometeorological variables lus convection. Mon Weather Rev 105(9):1171–1188. https://doi.
and atmospheric chemistry. Atmos Chem Phys 21(1):339–356. org/10.1175/1520-0493(1977)105%3C1171:DVODCC%3E2.0
https://doi.org/10.5194/acp-21-339-2021 .CO;2
Campos S, Mendes KR, da Silva LL, Mutti PR, Medeiros SS, Amorim Hervada-Sala C, Jarauta-Bragulat E (2004) A program to perform
LB, Bezerra BG (2019) Closure and partitioning of the energy Ward’s clustering method on several regionalized variables.
balance in a preserved area of a Brazilian seasonally dry tropical Comput Geosci 30(8):881–886. https://doi.org/10.1016/j.
forest. Agric for Meteorol 271:398–412. https://doi.org/10.1016/j. cageo.2004.07.003
agrformet.2019.03.018 Horel JD, Hahmann AN, Geisler JE (1989) An investiga-
Carvalho LM, Jones C, Liebmann B (2004) The South Atlantic con- tion of the annual cycle of convective activity over the
vergence zone: intensity, form, persistence, and relationships with tropical Americas. J Clim 2(11):1388–1403. https://doi.
intraseasonal to interannual activity and extreme rainfall. J Clim org/10.1175/1520-0442(1989)002%3C1388:AIOTAC%3E2.0
17(1):88–108 .CO;2
Cohen JC, Silva Dias MA, Nobre CA (1995) Environmental Itterly KF, Taylor PC, Dodson JB, Tawfik AB (2016) On the sensitiv-
conditions associated with amazonian squall lines: a case ity of the diurnal cycle in the Amazon to convective intensity.
study. Mon Weather Rev 123(11):3163–3174. https://doi. J Geophys Research: Atmos 121(14):8186–8208. https://doi.
org/10.1175/1520-0442(2004)017%3C0088:TSACZI%3E2.0 org/10.1002/2016JD025039
.CO;2 Janowiak JE, Kousky VE, Joyce RJ (2005) Diurnal cycle of precipita-
Cui Z, Zhang GJ, Wang Y, Xie S (2021) Understanding the roles of tion determined from the CMORPH high spatial and temporal
convective trigger functions in the diurnal cycle of precipita- resolution global precipitation analyses. J Geophys Research:
tion in the NCAR CAM5. J Clim 34(15):6473–6489. https://doi. Atmos 110:D23. https://doi.org/10.1029/2005JD006156
org/10.1175/JCLI-D-20-0699.1 Ketchen DJ, Shook CL (1996) The application of cluster analysis in
Da Rocha RP, Morales CA, Cuadra SV, Ambrizzi T (2009) Precipi- strategic management research: an analysis and critique. Strateg
tation diurnal cycle and summer climatology assessment over Manag J 17(6):441–458. https://doi.org/10.1002/(SICI)1097-
South America: an evaluation of Regional Climate Model version 0266(199606)17:6%3C441::AID-SMJ819%3E3.0.CO;2-G
3 simulations. J Geophys Research: Atmos 114:D10. https://doi. Khanna J, Medvigy D, Fisch G, de Araújo T, Neves TT (2018)
org/10.1029/2008JD010212 Regional hydroclimatic variability due to contemporary defor-
Daleu CL, Plant RS, Woolnough SJ, Stirling AJ, Harvey NJ (2020) estation in southern Amazonia and associated boundary layer
Memory properties in cloud-resolving simulations of the diurnal characteristics. J Geophys Research: Atmos 123(8):3993–4014.
cycle of deep convection. J Adv Model Earth Syst 12(8). https:// https://doi.org/10.1002/2017JD027888
doi.org/10.1029/2019MS001897 Kong X, Wang A, Bi X, Wei J, Li X (2023) The influence of differ-
De Souza EB, Kayano MT, Ambrizzi T (2005) Intraseasonal and sub- ent parameterizations on diurnal cycle of land precipitation in
monthly variability over the eastern Amazon and Northeast Brazil CAS-ESM. Atmos Res 282:106511. https://doi.org/10.1016/j.
during the autumn rainy season. Theoret Appl Climatol 81:177– atmosres.2022.106511
191. https://doi.org/10.1007/s00704-004-0081-4 Kousky VE (1980) Diurnal rainfall variation in northeast Bra-
dos Reis JS, Gonçalves WA, Mendes D (2021) Climatology of the zil. Mon Weather Rev 108(4):488–498. https://doi.
dynamic and thermodynamic features of upper tropospheric org/10.1175/1520-0493(1980)108%3C0488:DRVINB%3E2.0
cyclonic vortices in Northeast Brazil. Clim Dyn 57(11):3413– .CO;2
3431. https://doi.org/10.1007/s00382-021-05873-y Lapola DM, Martinelli LA, Peres CA, Ometto JP, Ferreira ME,
Durieux L, Machado LAT, Laurent H (2003) The impact of defor- Nobre CA, Vieira IC (2014) Pervasive transition of the Brazil-
estation on cloud cover over the Amazon arc of deforestation. ian land-use system. Nat Clim Change 4(1):27–35. https://doi.
Remote Sens Environ 86(1):132–140. https://doi.org/10.1016/ org/10.1038/nclimate2056
S0034-4257(03)00095-6 Lewis E, Fowler H, Alexander L, Dunn R, McClean F, Barbero R, Blen-
Durkee JD, Mote TL, Shepherd JM (2009) The contribution of meso- kinsop S (2019) GSDR: a global sub-daily rainfall dataset. J Clim
scale convective complexes to rainfall across subtropical South 32(15):4715–4729. https://doi.org/10.1175/JCLI-D-18-0143.1
America. J Clim 22(17):4590–4605. https://doi.org/10.1175/200 Lima GN, Rueda VOM (2018) The urban growth of the metropolitan
9JCLI2858.1 area of Sao Paulo and its impact on the climate. Weather Clim
Escobar GCJ, Reboita MS, Souza A (2019) Climatology of surface Extremes 21:17–26. https://doi.org/10.1016/j.wace.2018.05.002
baroclinic zones. Coast Brazil Atmósfera 32(2):129–141. https:// Lin L, Gettelman A, Xu Y, Wu C, Wang Z, Rosenbloom N, Dong W
doi.org/10.20937/atm.2019.32.02.04 (2019) CAM6 simulation of mean and extreme precipitation over
Evans JL, Braun A (2012) A climatology of subtropical cyclones Asia: sensitivity to upgraded physical parameterizations and
in the South Atlantic. J Clim 25(21):7328–7340. https://doi. higher horizontal resolution. Geosci Model Dev 12(8):3773–
org/10.1175/JCLI-D-11-00212.1 3793. https://doi.org/10.5194/gmd-12-3773-2019
Fisch G, Tota J, Machado LAT, Silva Dias MAFD, Da F, Lyra RF, Marengo JA, Soares WR, Saulo C, Nicolini M (2004) Climatol-
Nobre CA, Gash JHC (2004) The convective boundary layer over ogy of the low-level jet east of the Andes as derived from
pasture and forest in Amazonia. Theoret Appl Climatol 78:47–59. the NCEP–NCAR reanalyses: characteristics and tem-
https://doi.org/10.1007/s00704-004-0043-x poral variability. J Clim 17(12):2261–2280. https://doi.
Giles JA, Ruscica RC, Menéndez CG (2020) The diurnal cycle of org/10.1175/1520-0442(2004)017%3C2261:COTLJE%3E2.0
precipitation over South America represented by five gridded .CO;2
datasets. Int J Climatol 40(2):668–686. https://doi.org/10.1002/ Marengo JA, Liebmann B, Grimm AM, Misra V, Silva Dias PLD, Cav-
joc.6229 alcanti IFA, Alves LM (2012) Recent developments on the south

13
Content courtesy of Springer Nature, terms of use apply. Rights reserved.
Diurnal cycle of precipitation in Brazil 7825

American monsoon system. Int J Climatol 32(1):1–21. https:// past eight decades. Int J Climatol 41(8):4171–4186. https://doi.
doi.org/10.1002/joc.2254 org/10.1002/joc.7065
Mattos EV, Machado LA, Williams ER, Goodman SJ, Blakeslee RJ, Raia A, Cavalcanti IFA (2008) The life cycle of the south American
Bailey JC (2017) Electrification life cycle of incipient thunder- monsoon system. J Clim 21(23):6227–6246. https://doi.org/10.1
storms. J Geophys Research: Atmos 122(8):4670–4697. https:// 175/2008JCLI2249.1
doi.org/10.1002/2016JD025772 Rao VB, Franchito SH, Santo CM, Gan MA (2016) An update on
Meredith EP, Ulbrich U, Rust HW (2019) The diurnal nature of the rainfall characteristics of Brazil: seasonal variations and
future extreme precipitation intensification. Geophys Res Lett trends in 1979–2011. Int J Climatol 36(1):291–302. https://doi.
46(13):7680–7689. https://doi.org/10.1029/2019GL082385 org/10.1002/joc.4345
Mimmack GM, Mason SJ, Galpin JS (2001) Choice of distance matri- Reboita MS, Gan MA, Rocha RPD, Ambrizzi T (2010) Regimes de
ces in cluster analysis: defining regions. J Clim 14(12):2790– precipitação na América do sul: uma revisão bibliográfica. Revista
2797. https://doi.org/10.1175/1520-0442(2001)014%3C2790:C brasileira de meteorologia 25:185–204. https://doi.org/10.1590/
ODMIC%3E2.0.CO;2 S0102-77862010000200004
Mingoti SA (2005) Análise de Dados Através de Métodos de Estatística Reboita MS, Dutra LMM, Dias CG (2016) Diurnal cycle of pre-
- Multivariada - Uma Abordagem Aplicada. UFMG, 2005 cipitation simulated by RegCM4 over South America: pres-
Minobe S, Park JH, Virts KS (2020) Diurnal cycles of precipitation and ent and future scenarios. Climate Res 70(1):39–55. https://doi.
lightning in the tropics observed by TRMM3G68, GSMaP, LIS, org/10.3354/cr01416
and WWLLN. J Clim 33(10):4293–4313. https://doi.org/10.1175/ Reis L, Santos e Silva CM, Bezerra B, Mutti P, Spyrides MH, Silva
JCLI-D-19-0389.1 P, Andrade L (2020) Influence of climate variability on soybean
Montini TL, Jones C, Carvalho LM (2019) The south American yield in Matopiba, Brazil. Atmosphere 11(10):1130. https://doi.
low-level jet: a new climatology, variability, and changes. J org/10.3390/atmos11101130
Geophys Research: Atmos 124(3):1200–1218. https://doi. Rodrigues DT, Gonçalves WA, Spyrides MHC, Santos e, Silva CM
org/10.1029/2018JD029634 (2020) Spatial and temporal assessment of the extreme and daily
Moraes FD, Aquino FE, Mote TL, Durkee JD, Mattingly KS (2020) precipitation of the Tropical Rainfall Measuring Mission satellite
Atmospheric characteristics favorable for the development of in Northeast Brazil. Int J Remote Sens 41(2):549–572. https://doi.
mesoscale convective complexes in southern Brazil. Climate Res org/10.1080/01431161.2019.1643940
80(1):43–58. https://doi.org/10.3354/cr01595 Rodrigues DT, Gonçalves WA, Spyrides MHC, Andrade LDMB, de
Mutti PR, da Silva LL, Medeiros SDS, Dubreuil V, Mendes KR, Souza DO, de Araujo PAA (2021) … Santos e Silva, C. M. Prob-
Marques TV, Bezerra BG (2019) Basin scale rainfall-evapotrans- ability of occurrence of extreme precipitation events and natural
piration dynamics in a tropical semiarid environment during dry disasters in the city of Natal, Brazil. Urban Climate, 35:100753.
and wet years. Int J Appl Earth Obs Geoinf 75:29–43. https://doi. https://doi.org/10.1016/j.uclim.2020.100753
org/10.1016/j.jag.2018.10.007 Rodrigues DT, Gonçalves WA, Santos e Silva CM, Spyrides MHC,
Nesbitt SW, Zipser EJ (2003) The diurnal cycle of rainfall and Lúcio PS (2023) Imputation of precipitation data in north-
convective intensity according to three years of TRMM east Brazil. An Acad Bras Cienc 95:e20210737. https://doi.
measurements. J Clim 16(10):1456–1475. https://doi. org/10.1590/0001-3765202320210737
org/10.1175/1520-0442(2003)016%3C1456:TDCORA%3E2.0 Ruppert JH Jr, Klocke D (2019) The two diurnal modes of tropical
.CO;2 upward motion. Geophys Res Lett 46(5):2911–2921. https://doi.
Ogura Y, Takahashi T (1971) Numerical simulation of the life cycle of org/10.1029/2018GL081806
a thunderstorm cell. Mon Weather Rev 99(12):895–911. https:// Salio P, Nicolini M, Zipser EJ (2007) Mesoscale convective systems
doi.org/10.1175/1520-0493(1971)099%3C0895:NSOTLC%3E2 over southeastern South America and their relationship with the
.3.CO;2 south American low-level jet. Mon Weather Rev 135(4):1290–
Oliveira PD, Santos e Silva CM, Lima KC (2017) Climatology and 1309. https://doi.org/10.1175/MWR3305.1
trend analysis of extreme precipitation in subregions of North- Santos e, Silva CM (2013) Ciclo diário E semidiário de precipitação na
east Brazil. Theoret Appl Climatol 130:77–90. https://doi. costa norte do Brasil. Revista Brasileira De Meteorologia 28:34–
org/10.1007/s00704-016-1865-z 42. https://doi.org/10.1590/S0102-77862013000100004
Oliveira MI, Nascimento EL, Kannenberg C (2018) A new look Santos e, Silva CM, Gielow R, de Freitas SR (2009) Diurnal and semi-
at the identification of low-level jets in South America. Mon diurnal rainfall cycles during the rain season in SW Amazonia,
Weather Rev 146(7):2315–2334. https://doi.org/10.1175/ observed via rain gauges and estimated using S-band radar. Atmo-
MWR-D-17-0237.1 spheric Sci Lett 10(2):87–93. https://doi.org/10.1002/asl.214
Onumanyi AJ, Bello-Salau H, Adejo AO, Ohize HO, Oloyede MO, Santos e, Silva CM, de Freitas SR, Gielow R (2012) Numerical simu-
Paulson EN, Aibinu AM (2020) A discriminant analysis- lation of the diurnal cycle of rainfall in SW Amazon basin dur-
based automatic ordered statistics scheme for radar systems. ing the 1999 rainy season: the role of convective trigger function.
Phys Communication 43:101215. https://doi.org/10.1016/j. Theoret Appl Climatol 109:473–483. https://doi.org/10.1007/
phycom.2020.101215 s00704-011-0571-0
Pezzi LP, Quadro MF, Souza EB, Miller AJ, Rao VB, Rosa EB, Santos EB, Lucio PS, Santos e, Silva CM (2016) Estimating return
António JF (2023) Oceanic SACZ produces an abnormally wet periods for daily precipitation extreme events over the Brazil-
2021/2022 rainy season in South America. Sci Rep 13(1):1455. ian Amazon. Theoret Appl Climatol 126:585–595. https://doi.
https://doi.org/10.1038/s41598-023-28803-w org/10.1007/s00704-015-1605-9
Pichelli E, Coppola E, Sobolowski S, Ban N, Giorgi F, Stocchi P, Silva JDS, Cabral Júnior JB, Rodrigues DT, Silva FDDS (2023) Cli-
Vergara-Temprado J (2021) The first multi-model ensemble of matology and significant trends in air temperature in Alagoas,
regional climate simulations at kilometer-scale resolution part Northeast Brazil. Theoret Appl Climatol 151(3):1805–1824.
2: historical and future simulations of precipitation. Clim Dyn https://doi.org/10.1007/s00704-023-04358-8
56:3581–3602. https://doi.org/10.1007/s00382-021-05657-4 Silva Dias PL, Bonatti JP, Kousky VE (1987) Diurnally
Prado LF, Wainer I, Yokoyama E, Khodri M, Garnier J (2021) Changes forced tropical tropospheric circulation over South Amer-
in summer precipitation variability in central Brazil over the ica. Mon Weather Rev 115(8):1465–1478. https://doi.

13
Content courtesy of Springer Nature, terms of use apply. Rights reserved.
7826 C. M. Santos e Silva et al.

org/10.1175/1520-0493(1987)115%3C1465:DFTTCO%3E2.0 Wang J, Chagnon FJ, Williams ER, Betts AK, Renno NO, Machado
.CO;2 LA, Bras RL (2009) Impact of deforestation in the Amazon basin
Silva Dias MD, Rutledge S, Kabat P, Silva Dias PD, Nobre C, on cloud climatology. Proc Natl Acad Sci 106(10):3670–3674.
Fisch G, Gatti L (2002) Cloud and rain processes in a bio- https://doi.org/10.1073/pnas.0810156106
sphere-atmosphere interaction context in the Amazon region. Ward JH Jr (1963) Hierarchical grouping to optimize an objective
J Geophys Research: Atmos 107(D20):LBA–39. https://doi. function. J Am Stat Assoc 58(301):236–244. https://doi.org/10.10
org/10.1029/2001JD000335 80/01621459.1963.10500845
Siqueira RAD, Vila DA, Afonso JMDS (2021) The performance of the Watters D, Battaglia A, Allan RP (2021) The diurnal cycle of precipi-
diurnal cycle of precipitation from blended satellite techniques tation according to multiple decades of global satellite observa-
over Brazil. Remote Sens 13(4):734. https://doi.org/10.3390/ tions, three CMIP6 models, and the ECMWF reanalysis. J Clim
rs13040734 34(12):5063–5080. https://doi.org/10.1175/JCLI-D-20-0966.1
Sokal RR, Rohlf FJ (1962) The comparison of dendrograms by objective Webster PJ, Clayson CA, Curry JA (1996) Clouds, radiation, and
methods. Taxon 11(2):33–40. https://doi.org/10.2307/1217208 the diurnal cycle of sea surface temperature in the tropi-
Souza DCD, Oyama MD (2017) Breeze potential along the Brazil- cal western Pacific. J Clim 9(8):1712–1730. https://doi.
ian northern and northeastern coast. J Aerosp Technol Manage org/10.1175/1520-0442(1996)009%3C1712:CRATDC%3E2.0
9:368–378. https://doi.org/10.5028/jatm.v9i3.787 .CO;2
Spangler KR, Lynch AH, Spera SA (2017) Precipitation drivers of Wimmer D, Buenrostro Mazon S, Manninen HE, Kangasluoma J,
cropping frequency in the Brazilian Cerrado: evidence and impli- Franchin A, Nieminen T, Petäjä T (2018) Ground-based obser-
cations for decision-making. Weather Clim Soc 9(2):201–213. vation of clusters and nucleation-mode particles in the Amazon.
https://doi.org/10.1175/WCAS-D-16-0024.1 Atmos Chem Phys 18(17):13245–13264. https://doi.org/10.5194/
Sui CH, Lau KM, Takayabu YN, Short DA (1997) Diurnal acp-18-13245-2018
variations in tropical oceanic cumulus convection dur- Yan Y, Zhang L, Yu Y, Chen C, Xi J, Chai F (2021) Rectification of the
ing TOGA COARE. J Atmos Sci 54(5):639–655. https://doi. intraseasonal SST variability by the diurnal cycle of SST revealed
org/10.1175/1520-0469(1997)054%3C0639:DVITOC%3E2.0 by the global tropical moored buoy array. Geophys Res Lett
.CO;2 48(1). https://doi.org/10.1029/2020GL090913. e2020GL090913
Tenório RS, Moraes MCDS, Kwon BH (2010) Raindrop distribution Yang S, Smith EA (2006) Mechanisms for diurnal variability of global
in the Eastern Coast of Northeastern Brazil using disdrometer tropical rainfall observed from TRMM. J Clim 19(20):5190–
data. Revista Brasileira De Meteorologia 25:415–426. https://doi. 5226. https://doi.org/10.1175/JCLI3883.1
org/10.1590/S0102-77862010000400001 Yang S, Kuo KS, Smith EA (2008) Persistent nature of secondary
Tripoli GJ, Cotton WR (1989) Numerical study of an observed oro- diurnal modes of precipitation over oceanic and continental
genic mesoscale convective system. Part 1: simulated genesis and regimes. J Clim 21(16):4115–4131. https://doi.org/10.1175/200
comparison with observations. Mon Weather Rev 117(2):273– 8JCLI2140.1
304. https://doi.org/10.1175/1520-0493(1989)117%3C0273:NS
OAOO%3E2.0.CO;2 Publisher’s Note Springer Nature remains neutral with regard to juris-
Wallace JM (1975) Diurnal variations in precipitation and dictional claims in published maps and institutional affiliations.
thunderstorm frequency over the conterminous United
States. Mon Weather Rev 103(5):406–419. https://doi. Springer Nature or its licensor (e.g. a society or other partner) holds
org/10.1175/1520-0493(1975)103%3C0406:DVIPAT%3E2.0 exclusive rights to this article under a publishing agreement with the
.CO;2 author(s) or other rightsholder(s); author self-archiving of the accepted
Wang J, Bras RL, Eltahir EA (2000) The impact of observed defor- manuscript version of this article is solely governed by the terms of
estation on the mesoscale distribution of rainfall and clouds such publishing agreement and applicable law.
in Amazonia. J Hydrometeorol 1(3):267–286. https://doi.
org/10.1175/1525-7541(2000)001%3C0267:TIOODO%3E2.0
.CO;2

13
Content courtesy of Springer Nature, terms of use apply. Rights reserved.
Terms and Conditions
Springer Nature journal content, brought to you courtesy of Springer Nature Customer Service Center GmbH (“Springer Nature”).
Springer Nature supports a reasonable amount of sharing of research papers by authors, subscribers and authorised users (“Users”), for small-
scale personal, non-commercial use provided that all copyright, trade and service marks and other proprietary notices are maintained. By
accessing, sharing, receiving or otherwise using the Springer Nature journal content you agree to these terms of use (“Terms”). For these
purposes, Springer Nature considers academic use (by researchers and students) to be non-commercial.
These Terms are supplementary and will apply in addition to any applicable website terms and conditions, a relevant site licence or a personal
subscription. These Terms will prevail over any conflict or ambiguity with regards to the relevant terms, a site licence or a personal subscription
(to the extent of the conflict or ambiguity only). For Creative Commons-licensed articles, the terms of the Creative Commons license used will
apply.
We collect and use personal data to provide access to the Springer Nature journal content. We may also use these personal data internally within
ResearchGate and Springer Nature and as agreed share it, in an anonymised way, for purposes of tracking, analysis and reporting. We will not
otherwise disclose your personal data outside the ResearchGate or the Springer Nature group of companies unless we have your permission as
detailed in the Privacy Policy.
While Users may use the Springer Nature journal content for small scale, personal non-commercial use, it is important to note that Users may
not:

1. use such content for the purpose of providing other users with access on a regular or large scale basis or as a means to circumvent access
control;
2. use such content where to do so would be considered a criminal or statutory offence in any jurisdiction, or gives rise to civil liability, or is
otherwise unlawful;
3. falsely or misleadingly imply or suggest endorsement, approval , sponsorship, or association unless explicitly agreed to by Springer Nature in
writing;
4. use bots or other automated methods to access the content or redirect messages
5. override any security feature or exclusionary protocol; or
6. share the content in order to create substitute for Springer Nature products or services or a systematic database of Springer Nature journal
content.
In line with the restriction against commercial use, Springer Nature does not permit the creation of a product or service that creates revenue,
royalties, rent or income from our content or its inclusion as part of a paid for service or for other commercial gain. Springer Nature journal
content cannot be used for inter-library loans and librarians may not upload Springer Nature journal content on a large scale into their, or any
other, institutional repository.
These terms of use are reviewed regularly and may be amended at any time. Springer Nature is not obligated to publish any information or
content on this website and may remove it or features or functionality at our sole discretion, at any time with or without notice. Springer Nature
may revoke this licence to you at any time and remove access to any copies of the Springer Nature journal content which have been saved.
To the fullest extent permitted by law, Springer Nature makes no warranties, representations or guarantees to Users, either express or implied
with respect to the Springer nature journal content and all parties disclaim and waive any implied warranties or warranties imposed by law,
including merchantability or fitness for any particular purpose.
Please note that these rights do not automatically extend to content, data or other material published by Springer Nature that may be licensed
from third parties.
If you would like to use or distribute our Springer Nature journal content to a wider audience or on a regular basis or in any other manner not
expressly permitted by these Terms, please contact Springer Nature at

[email protected]