\makeatletter
\@ifundefined{HCode}
{\documentclass[screen,CRGEOS,Unicode,Thematic]{cedram}
\usepackage[square,comma,authoryear]{natbib}
\renewcommand\bibsection{\section*{\refname}}
\newenvironment{noXML}{}{}
\def\thead{\noalign{\relax}\hline}
\def\endthead{\noalign{\relax}\hline}
\def\tbody{}%\noalign{\relax}\hline}
\def\tsup#1{$^{{#1}}$}
\def\tsub#1{$_{{#1}}$}
\def\ndash{\text{--}}
\def\og{\guillemotleft}
\def\fg{\guillemotright}
\def\0{\phantom{0}}
\def\xsection#1{}
\makeatletter
\g@addto@macro{\UrlBreaks}{\UrlOrds}
\gappto{\UrlBreaks}{\UrlOrds}
\RequirePackage{etoolbox}
\usepackage{upgreek}
\usepackage[figuresright]{rotating}
\def\jobid{crgeos20230749}
%\graphicspath{{/tmp/\jobid_figs/web/}}
\graphicspath{{./figures/}}
\newcounter{runlevel}
\usepackage{multirow}
\let\MakeYrStrItalic\relax
\def\refinput#1{}
\let\ubreak\break
\csdef{Seqnsplit}{\\}
\def\botline{\\\hline}
\def\back#1{}
\def\rmmu{\upmu}
\def\inlinefig#1{\includegraphics{#1}}
\newenvironment{xxcenter}{\begin{center}}{\end{center}}
\DOI{10.5802/crgeos.246}
\datereceived{2023-08-25}
\daterevised{2023-11-21}
\dateaccepted{2023-11-21}
\ItHasTeXPublished
}
{
\PassOptionsToPackage{authoryear}{natbib}
\documentclass[crgeos]{article}
\usepackage[T1]{fontenc}
\def\CDRdoi{10.5802/crgeos.246}
\let\ubreak\relax
\makeatletter
\def\CDRsupplementaryTwotypes#1#2{}
\newenvironment{xxcenter}{}{}
}
\makeatother

\usepackage{upgreek}

\dateposted{2023-12-18}
\begin{document}

\begin{noXML}

%\makeatletter
%\def\TITREspecial{\relax}
%\def\cdr@specialtitle@english{Tribute to Jean Dercourt}
%%\def\cdr@specialtitle@french{Hommage \`a Jean Dercourt}
%\makeatother

\title{The Br\'{e}cy depocenter as part of a new northern Massif Central
Carboniferous--Permian Basin~(France)}

\alttitle{Le d\'{e}pocentre de Br\'{e}cy appartient \`{a} un nouveau
bassin carbonif\`{e}re--permien au nord du Massif central (France)}

\author{\firstname{Laurent} \lastname{Beccaletto}\CDRorcid{0000-0003-2132-1738}\IsCorresp}
\address{BRGM, 45060 Orl\'{e}ans, France}
\email[L. Beccaletto]{l.beccaletto@brgm.fr}

\author{\firstname{Sylvie} \lastname{Bourquin}\CDRorcid{0000-0002-2802-9548}}
\address{Univ Rennes, CNRS, G\'{e}osciences Rennes - UMR 6118, 35000
Rennes, France}
\email[S. Bourquin]{sylvie.bourquin@univ-rennes1.fr}

\begin{abstract}
The reinterpretation of deep wells and the reprocessing and
interpretation of 115 km of industrial seismic lines can be used to
update the geometry, depositional environments and tectonic evolution
of the Carboniferous--Permian Br\'{e}cy depocenter (southwest Paris
Basin). The present-day geometry of the Br\'{e}cy depocenter is
controlled by several eastward dipping normal faults, some of which
possibly connected to deep detachments that were active during the late
Carboniferous--Permian history. It is estimated that the maximum
thickness of the Br\'{e}cy depocenter is 3900~m. The filling provides
evidence for a thick late Carboniferous--lower Permian syn-rift stage
overlain by a thin post-rift stage, probably similar to the tectonic
evolution of the northeastward basins in the Lorraine region of France
and Germany and thus attributed to the lower-middle Permian. The facies
generally characterize lake environments, with occurrences of sediment
supply attributed to fluvial, alluvial fan and delta fan deposits. They
mainly display a retrogradational--progradational pattern during the
syn-rift stage, and a retrogradational pattern during the post-rift
stage. The Br\'{e}cy area was part of a larger late Variscan basin
during the latest Carboniferous--early Permian times in the northern
Massif Central region.
\end{abstract}

\begin{altabstract}
La r\'{e}interpr\'{e}tation de puits profonds et le retraitement et
l'interpr\'{e}tation de 115 km de lignes sismiques industrielles
permettent de mettre \`{a} jour la g\'{e}om\'{e}trie, les
environnements de d\'{e}p\^{o}t et l'\'{e}volution tectonique du
d\'{e}pocentre carbonif\`{e}re--permien de Br\'{e}cy (sud-ouest du
Bassin de Paris). La g\'{e}om\'{e}trie actuelle du d\'{e}pocentre de
Br\'{e}cy est contr\^{o}l\'{e}e par plusieurs failles normales \`{a}
pendage vers l'est, dont certaines sont possiblement connect\'{e}es
\`{a} des d\'{e}tachements profonds actifs \`{a} la fin du
Carbonif\`{e}re et au Permien. On estime que l'\'{e}paisseur maximale
du d\'{e}pocentre de Br\'{e}cy est de 3900~m. Le remplissage
s\'{e}dimentaire montre l'existence d'un \'{e}pais stade syn-rift \`{a}
la fin du Carbonif\`{e}re et au Permien inf\'{e}rieur, recouvert par un
stade post-rift peu \'{e}pais. Cette \'{e}volution tectonique est
probablement similaire \`{a} celle connue vers le nord-est en Lorraine
et Allemagne, attribu\'{e}e au Permien inf\'{e}rieur-moyen. Les
faci\`{e}s t\'{e}moignent g\'{e}n\'{e}ralement d'environnements
lacustres, avec des apports s\'{e}dimentaires attribu\'{e}s \`{a} des
d\'{e}p\^{o}ts fluviaux, de c\^{o}nes alluviaux et de c\^{o}nes
delta\"{i}ques. Ils expriment un syst\`{e}me
r\'{e}trogradant-progradant pendant la phase syn-rift, et un
syst\`{e}me r\'{e}trogradant pendant la phase post-rift. A la fin du
Carbonif\`{e}re et au d\'{e}but du Permien, la r\'{e}gion de Br\'{e}cy
faisait partie d'un bassin tardi-varisque plus vaste situ\'{e} au nord
du Massif central.
\end{altabstract}

\keywords{\kwd{Late Carboniferous}
\kwd{Permian}
\kwd{Paris Basin}
\kwd{Deep well}
\kwd{Seismic interpretation}
\kwd{Regional correlation}}

\altkeywords{\kwd{Carbonif\`{e}re sup\'{e}rieur}
\kwd{Permien}
\kwd{Bassin de Paris}
\kwd{Puits profond}
\kwd{Interpr\'{e}tation sismique}
\kwd{Corr\'{e}lation r\'{e}gionale}}

\maketitle

\vspace*{3pt plus 1pt minus 1pt}
\twocolumngrid

\end{noXML}

\section{Introduction}

The late Carboniferous--Permian is a period of large-scale geodynamic
reorganization, (i) with the final accretion of Pangea
(Variscides/Mauretanides/{\ubreak}Alleghenides orogeneses) and the beginning of
its breakup (NeoTethys opening)  
\citep[e.g.,][]{DomeierTorsvik2014,ScoteseLangford1995,Stampflietal2013},
coeval with (ii) the acme of the Late Paleozoic Ice Age (LPIA),
constituting a turning point in the climate regimes of the Paleozoic 
\citep[e.g.,][]{Gastaldoetal1996,Montanezetal2007}.

From the end of the Carboniferous onwards, the over-thickened and hot
Variscan crust collapsed [late-orogenic collapse;
\citealp{Burgetal1994,Faure1995,Malavieilleetal1990,MenardMolnar1988,VanDenDriesscheBrun1992}], 
triggering the rise of high-grade metamorphic domes along low-angle
detachment faults, and the development of half-graben or pull-apart
multi-directional intramountain coal basins 
\citep{CostaRey1995,Gardienetal2022,Malavieille1993,McCannetal2008a,McCannetal2008b, MenardMolnar1988,VanDenDriesscheBrun1989}.

These late orogenic Carboniferous--Permian basins (hereafter LOCPB)
therefore widely developed around 300 Ma in the internal parts of the
belt south of the northern Variscan Front. They crop out in several
limited locations in and around the Variscan basement of western Europe
(Pyrenees, Massif Central, Brittany, Vosges-Black Forest, Alps, Harz),
in close genetic relationships with major Variscan tectonic structures 
(e.g., Sillon Houiller, South-Hunsr\"{u}ck Fault; Figure~\ref{fig1} and
related references).\looseness=-1

\begin{figure*}
\vspace*{2pt}
\includegraphics{fig01}
\caption{\label{fig1}Map of NW European outcropping and subsurface
LOCPB; compiled from \citet{Beccalettoetal2015,BRGM1984, BRGM1989,
BRGM2003,Delmasetal2002,Gastetal2010, SchaferKorsch1998,
SchneiderRomer2010, SchneiderScholze2018}. 
\textit{Numbers}---French outcropping LOCPB: 1:~Carentan/Littry; 2:
St-Pierre-la-Cour; 3: Dou\'{e}-la-Fontaine; 4: Sillon Vend\'{e}en,
Chantonnay, Faymoreau; 5:~Vill\'{e}; 6: St-Di\'{e}; 7:
Ronchamp-Giromagny; 8: Autun; 9: Decize La Machine; 10: Blanzy-Le
Creusot; 11:~Bert; 12: Aumance; 13: Commentry, Doyet, Deneuille; 14:
Ahun; 15: Bosmoreau-les-Mines; 16: Bassins du Sillon-Houiller; 17:
Brive; 18: Brassac, Brioude; 19: St-Etienne; 20: Prades-Jaujac; 21:
Al\`{e}s; 22: Graissessac, Lod\`{e}ve; 23: St-Affrique; 24: Figeac,
Decazeville, Rodez; 25: Quercy-Albigeois, La Gr\'{e}signe; 26: La
Rhune-Bidarray; 27: Ossau; 28: Var; 29: Barrot, Argentera; 30: Alpes
zones externes; 31: Alpes zones internes. \textit{Italic
letters}---Other West European LOCPB  A: Albersweiler; BI:
Bifertengr\"{a}tli; BR: Breisgau; EN: Entlebuch; FR: Frankenberg Bay
(Naab---Weiden); IF: Ilfeld; K: Kraichgau (Baden-Baden); LO:
Lombardian; MF: Manx-Furness; NSW: North Switzerland; S: Saale (Halle);
SCH: Schramberg; SNB: Saar-Nahe; RS: Ries-Salzach; S: Sudetic;  SD:
Salvan-Dorenaz; ST: Stockheim; TF: Thuringian Forest; VFD: Variscan
Foredeep; W: Wetterau; WE: Wessex;WI: Wittlich Graben. Blue
letters---\textit{Basement rocks}  A: Ardennes; Al: Alps; BF: Black
Forest; BM: Bohemian Massif; Br: Bretagne; H: Harz; MC: Massif Central;
Py: Pyrenees; SG: Schiefere Gebirge; V: Vosges. Sillon Houiller fault:
SH; South Hunsr\"{u}ck Fault: SHF. Arpheuilles-1 well: Arph.}
\end{figure*}

They were filled with siliciclastic volcanoclastic continental
material---from alluvial to lacustrine through fluvial
environments---at intertropical latitudes 
\citep{Donsimoni1990,Gand2003,Mercuzotetal2021,Mercuzotetal2022, 
Schafer2011,SchneiderRomer2010,Soreghanetal2020}. LOCPB were also
accompanied by widespread intrusive and extrusive magmatic activity
with a crustal or mantle origin with mostly felsic and rare mafic
signatures  \citep{Neumannetal2004,Timmerman2004, 
McCannetal2006}. Apart from scientific research, LOCPB were
extensively studied for their resources such as oil, gas, coal, uranium
or other ore deposits 
\citep{Bouchotetal1997, Bouchotetal2005,
Coureletal1986,Delmasetal2002,Dilletal1991,Mascle1990}, 
and more recently for geothermal \citep{Aretzetal2016} or natural
helium \citep{Hauvilleetal2021} purposes.


In map view, they occur as small isolated and disconnected ``basins''
with incomplete sedimentary successions and numerous sedimentation and
erosional gaps  \citep[e.g.,][]{Schneideretal2020}. Their present-day
area does not reflect their initial extent and thickness, which can be
explored by studying their subsurface prolongation beneath the
Meso-Cenozoic sedimentary covers 
\citep{Beccalettoetal2015,Mercuzotetal2021,SchneiderScholze2018,Ziegler1990}. 
These basins usually reach thicknesses of several kilometers like, for
instance, in the southern French Massif Central  [e.g., Lod\`{e}ve,
Carmaux---La Gr\'{e}sine; \citealp{BRGM1989}] or central Europe [e.g.,
Saar-Nahe and Thuringian Forest basins; \citealp{SchneiderRomer2010}].
Generally speaking, the ages of the LOCPB broadly range from Gzhelian
to Guadalupian encompassing the Stephanian, Autunian, lower Rotliegend
and part of the upper Rotliegend western European stages [ca. 303 Ma to
ca. 270 Ma; e.g., \citealp{Ducassouetal2019,Micheletal2015,Lutzneretal2020,
Oplustiletal2016,Pellenardetal2017,Poujoletal2023,Voigtetal2022}].

One recurrent question that arises from the study of the LOCPB is based
on sedimentological, structural and temporal criteria: what are their
regional correlations and extents? The answer to this question calls
for the recognition of reference basins with enough useful data to make
correlations. On one hand, recent advances in their age calibration
using radiochronological methods have led to improved inter-basin
comparisons  \citep[e.g.,][]{Ducassouetal2019,Mercuzotetal2023, 
Oplustiletal2016,Pellenardetal2017,Schneideretal2020,Voigtetal2022}.
On the other hand, fewer significant works have been carried out
recently on their structural pattern and tectonic evolution 
\citep[][and references therein]{Beccalettoetal2015}, and therefore
the poor understanding of their tectonic framework limits their
comparison at local or regional scales.

One good way to discuss the tectono-sedimentary history of hidden
subsurface LOCBP is to use seismic data to look for them under their
Meso-Cenozoic sedimentary cover given that they may be considered as
fossil basins with preserved pre-Triassic depositional and structural
patterns. We present new results from the interpretation of industrial
seismic lines spanning 115 kilometers, while targeting the Br\'{e}cy
depocenter in the southwest Paris Basin, which has recently been
reprocessed within the framework of an International Continental
Scientific Drilling Program proposal [ICDP Deepdust project;
\citealp{Soreghanetal2020}]. First, we aim to discuss the structural
features, thickness and tectonic evolution of the Br\'{e}cy depocenter
and the related sedimentary filling using seismic and well data. Last,
we compare and propose correlations between the Br\'{e}cy depocenter
and other Carboniferous--Permian deposits in the northern Massif
Central, as well as other places in France and Germany.

\section{Geological setting of the Br\'{e}cy depocenter}

The Br\'{e}cy depocenter is one of the three several-kilometer-thick
depocenters recently revealed by the interpretation of reprocessed
vintage seismic lines in the southwestern part of the Paris Basin
[together with the Contres and Arpheuilles depocenters; Figures~\ref{fig1},~\ref{fig2};
\citealp{Beccalettoetal2015}]. It has a roughly elongated shape
striking N30 parallel to the northern trend of the Sillon-Houiller
fault zone. The opening of the basin was thought to be controlled by
the activity of several N030-trending normal faults accommodating the
deposition of the Stephanian-Permian deposits which have estimated
thicknesses up to 3000~m.

\begin{figure*}
\vspace*{2pt}
\includegraphics{fig02}
\vspace*{2pt}
\caption{\label{fig2}Geological map of the study area depicting the
Br\'{e}cy depocenter and adjacent LOCPB, the location of the
reprocessed seismic lines, and deep wells. The geological background is
from the 1:1,000,000 geological map of France \citep{BRGM2003}.}
\vspace*{3.5pt}
\end{figure*}


The age of the Carboniferous--Permian series in the Br\'{e}cy area comes
from the description of the Bertray-1, Br\'{e}cy-1 and
Saint-Georges-sur-Moulon-1 well cuttings (from drilling reports;
hereafter called the BTY1, BRC1 and SGS1 wells, respectively;
Figure~\ref{fig2}), and from palynological data from the core of the
scientific Couy-1 well  [\citealp{Orszag-Sperberetal1992}; hereafter
COU1; Figure~\ref{fig2}].

In the COU1 well, the sedimentary succession, attributed to the Permian
based on the presence of several Darwinulacea ostracods 
\citep{Orszag-Sperberetal1992}, is composed of approximately 250 m of
mainly red silty-clay sediments with sandstones and few conglomerates
at the base. This succession describes a general trend from alluvial
fan to shallow lake deposits; neither tonstein (volcanic ashes
transformed into clay minerals) nor palynological data exist to shorten
the age range \citep{Juncaletal2018}. In this area, the Permian
deposits end with the Triassic unconformity, overlain by middle Anisian
deposits [Early and early-Middle Triassic gap;
\citealp{Bourquinetal2006,Juncaletal2018}].

In BTY1, from 2963 m to 2842 m, the sedimentary succession is mainly
composed of black-gray silty-clay interbedded with detrital coal,
sandstones and conglomerates. The age is attributed to the late
Carboniferous based on: gastropods and ostracods from the
\textit{Carbonita} genus found in the calcareous level at the base,
bisaccate pollens at 2963~m and the presence of \textit{Lycospora
pellucida} and \textit{Florinites junior} pollen at 2850~m 
\citep{Bertray11987}.

In BRC1, below 1806 m, the facies are more carbonated with
argillaceous-dolomitic limestones interbedded in fine-grained
gray-green argillaceous sandstones. Between 1802 m and 1806 m, plant
fragments suggesting a Carboniferous age  \citep{Brecy11966} have been
found in micaceous silty-claystone.\looseness=1

The drilling of the SGS1 well has reached a Paleozoic succession from
1639 to 1878.3 m attributed to the Cambro-Silurian by comparison with
known facies in outcrops in the Massif Central 
\citep{Georges-sur-Moulon11964}.

In COU1, the basement, composed of metamorphic rocks, starts at a depth
of 941.65 m. It is overlain by a highly tectonized Paleozoic
sedimentary series \citep{Lorenzetal1987} until 925.35 m. The Paleozoic
upper part is composed of volcanoclastic deposits attributed to the
Stephanian \citep{Chantraineetal1992}; trachy-andesites sampled at the
bottom of the Stephanian unit (925.35 m and 941.65 m) yielded an Ar--Ar
plateau age of 301.6 ${\pm}$ 6.3 Ma \citep{CostaMaluski1988}. The
Stephanian volcanoclastic unit is overlain by Permian fluvio-lacustrine
deposits above an angular unconformity \citep{Lorenzetal1987}.

\section{Material and methods}

\subsection{Seismic data reprocessing}

The present study is based on the reprocessing of five seismic lines
representing approximately 115 km, acquired by the ESSOREP oil company
in 1984 and 1985 (CHER survey, denoted as CH; Figure~\ref{fig2}). This
new dataset completes the ca. 200 km of vintage industrial seismic
reflection profiles that were reprocessed and interpreted in the
Br\'{e}cy area in the initial study of \citet{Beccalettoetal2015}.

Given the acquisition parameters used in these exploration surveys,
mainly the frequency bandwidth of the seismic source and the recording
length (up to 4 s TWT---Two-Way-Time), the estimated depth of
investigation is roughly 7--8 km, with a vertical resolution of
approximately 25 m and 30--35 m respectively in the shallow and deepest
part of the lines. These vintage seismic data have been reprocessed
using modern Pre-Stack time migration (PSTM) methods and algorithms,
thereby significantly improving the quality of the resulting seismic
section and providing enhanced descriptions of the geological
structures  \citep{Beccalettoetal2011,Beccalettoetal2015}. Efforts were
focused on three key steps that were repeated several times throughout
the processing unit: (a) computing primary and residual static
corrections in order to remove the topographic and velocity effects of
the superficial rock layer, strongly affecting the seismic signal; (b)
detailed velocity analysis; and (c)~\mbox{various} methods of organized and
random noise attenuation. Pre-stack time migration enhanced the details
of the structural features and completed this reprocessing unit before
stacking the data [cf. \citealp{Laurentetal2021} for another example of
seismic reprocessing and interpretation in Carboniferous basins].

\subsection{Seismic interpretation and well-data}

In a first step, the seismic facies and reflector geometries were
interpreted by using nearby wells to calibrate and tie the seismic
lines for the main horizons. Unlike the previous regional study of
\citet{Beccalettoetal2015}, the BTY1, BRC1 and COU1 wells are located
directly on some of the newly reprocessed seismic lines (resp. CH41,
CH22 and CH34; Figure~\ref{fig2}), thereby preventing any lateral
projection bias. Once the targeted horizons were identified near the
BTY1 and BRC1 wells (deepest wells), they were correlated step by step,
from line to line, by comparing the seismic facies using the Gverse
Geophysics module of the Gverse suite ({\textcopyright} Landmark), and
by checking their 3D structural consistency (using the 3D viewer of the
interpretation software). Steps two and three correspond to the
building of the structural scheme and the thickness map of the
Br\'{e}cy depocenter, respectively. In a fourth step, the electrofacies
and depositional environments are described within the framework of the
seismic interpretation by using subsurface well data such as: (i) the
description of the cuttings from reports and well data (gamma Ray (GR),
sonic or neutron, and resistivity) for the BRC1, BTY1, and SGS1 wells;
and (ii) core description and complete well log data (GR, sonic,
resistivity, density, photo electric factor (Pef), and neutron) for the
COU1 well \citep{Juncaletal2018}. In~a last step, these four wells were
correlated to give a lithostratigraphic and sequential framework to the
Br\'{e}cy area.

\section{Results}

\subsection{Seismic interpretation, targeted horizons and seismic facies}\label{sec4.1}

Figures~\ref{fig3} and~\ref{fig4} display four examples of investigated
seismic profiles in an uninterpreted and interpreted form. The main
seismic features and units are discussed below from bottom to top.


\begin{figure*}
\includegraphics{fig03}
\caption{\label{fig3}W--E composite line CH41-20-2 and line CH24,
interpreted and uninterpreted, depicting the main structural features
of the Br\'{e}cy depocenter; blue arrows: Meso-Cenozoic faults, black
arrows Carboniferous--Permian and Meso-Cenozoic faults; see text for
details.}
\end{figure*}

\begin{figure*}
\includegraphics{fig04}
\caption{\label{fig4}N--S line CH3 and line CH21, interpreted and
uninterpreted, depicting the main structural features of the Br\'{e}cy
depocentre; blue arrows: Meso-Cenozoic faults, black arrows
Carboniferous--Permian and Meso-Cenozoic faults; see text for details.}
\end{figure*}

The base of the late Carboniferous--Permian sedimentary fill (Base Basin
Unconformity---BBU) corresponds to the top of the pre-Stephanian
substratum. The four wells in the Br\'{e}cy area do not reach the
substratum, however it is cut across by fifteen wells in the westward
Contres and Arpheuilles depocenters. There, the substratum could be
sedimentary, metamorphic or plutonic \citep{Beccalettoetal2015}. In
most places, the top of the pre-Stephanian substratum matches the top
of the acoustic basement, where the seismic signal becomes chaotic.
However, some sets of reflectors, belonging to pre-Stephanian units,
still appear locally below the BBU (line CH3, Figure~\ref{fig4}).
\looseness=-1

The end-Carboniferous--Permian deposits s.s. correspond to the seismic
facies observed between the Top Paleozoic Unit (TPU) and the BBU. In
general, they are stratified with an overall wedge-shape geometry
slightly thinning eastward, and continuous reflectors varying from low
to high amplitudes and medium to high frequencies; chaotic or
semi-transparent facies are also possible. The topmost preserved
Permian beds display toplap geometries below the TPU (Figures~\ref{fig3} and~\ref{fig4}).
However, in closer detail, it is possible to distinguish three seismic
sequences:

\begin{enumerate}[(1)]
\item A \textit{lower seismic sequence} with thicknesses less than 0.5 s
TWT made of low-frequency/(very) high-amplitude, continuous to
discontinuous reflectors; in the BTY1 well, these reflectors are
correlated with Stephanian conglomeratic and coal beds, as seen in the
Arpheuilles-1 well  \citep{Beccalettoetal2015};

\item a much thicker \textit{intermediate} \textit{seismic sequence} (up
to 2 s TWT) with lateral thickness variations in units A, B and C; in
particular, note the thickening of unit A toward the center of the
basin. These three units were deposited during normal fault activity
and, as a whole, represent the syn-rift stage of the Br\'{e}cy
depocenter (cf. Section~\ref{sec5.1});

\item a thinner \textit{upper seismic sequence}, (less than 0.25 s TWT),
made up of unit D, located just below the TPU; this unit does not
display any thickness variations related to fault activity. The seismic
facies are more transparent here than in the lower and intermediate
units, certainly due to less contrasting lithologies (cf. Section~\ref{sec4.1}).
The horizons of the \textit{intermediate seismic sequence} display a
toplap geometry beneath the base of the \textit{upper seismic
sequence}, indicating an erosional phase in between. This \textit{upper
seismic sequence} represents the post-rift stage of the Br\'{e}cy
depocenter (cf. Section~\ref{sec5.1}).
\end{enumerate}

The Mesozoic sedimentary cover of the Paris Basin unconformably
overlies the Permian Br\'{e}cy deposits and consists of parallel to
sub-parallel continuous reflectors with medium- to high-frequencies and
alternating medium- to high-amplitudes related to lithological
variations within the Triassic and Jurassic deposits. The reflectors
gently dip northward towards the center of the Paris Basin.

\subsection{Fault patterns and structural schemes}

All seismic lines display normal faults, which are the norm in the
Br\'{e}cy depocenter; the vast majority of the faults developed during
the Meso-Cenozoic, as suggested by the equal down-throw all along the
fault plane both across the Carboniferous--Permian and Mesozoic strata
(blue arrows in Figures~\ref{fig3} and~\ref{fig4}). Some faults
characterized by different thicknesses of Carboniferous--Permian
deposits on both sides are expected to have controlled the late
Carboniferous--Permian filling. The seismic lines do not clearly image
the prolongation of these normal faults through the base of the basin
downward to the acoustic basement (Figures~\ref{fig3},~\ref{fig4}).
Deeply seated reflectors located in the prolongation of eastward
dipping normal faults may correspond to low angle normal faults
(detachments; Figures~\ref{fig3},~\ref{fig4}; cf.
Section~\ref{sec5.1}). These initial Carboniferous--Permian faults also
display Meso-Cenozoic activity, as shown by the down-throw of the TPU
and overlying Mesozoic strata (black arrows in
Figures~\ref{fig3},~\ref{fig4}).

The reprocessing and interpretation of the additional seismic lines
make it possible to build a detailed structural scheme for the
Br\'{e}cy depocenter, greatly improving the spatial resolution of the
structural pattern drawn by \citet{Beccalettoetal2015}. The present
study leads to the recognition of seven ``major'' faults (i.e., with a
large cartographic extent and usually cutting through several seismic
lines), and eleven secondary faults (i.e., cutting through one or two
seismic lines to a lower extent; Figure~\ref{fig5}). Most of these
faults display an eastward dip, and only the antithetic fault has a
westward dip. Their orientation varies between 20 and 40\textdegree N
with an average value of approximately 30\textdegree N. As mentioned
above, the structural pattern is actually a Meso-Cenozoic scheme, as
all the faults---including those active during the
Carboniferous--Permian---were active during the Mesozoic. We do not
exclude the occurrence of smaller scale transverse faults  (e.g., with
N80 to N180 trends), but the numerous N30-trending faults may mask them
and/or the density of the seismic dataset does not allow to discern
them.


\begin{figure*}
\vspace*{-3pt}
\includegraphics{fig05}
\vspace*{-3pt}
\caption{\label{fig5}\fontsize{9.9}{12}\selectfont 
Structural scheme of the Br\'{e}cy depocenter,
depicting the major/minor faults, those with a Meso-Cenozoic tectonic
activity or a Permo-Carboniferous and Meso-Cenozoic tectonic activity,
and deep detachments; the geological background is from the 1:1,000,000
geological map of France \citep{BRGM2003}.}
\vspace*{-3pt}
\end{figure*}

\subsection{Thickness maps of the Br\'{e}cy depocenter}

The new seismic dataset is used to accurately estimate the thickness of
the Br\'{e}cy depocenter (Figure~\ref{fig6}). Thickness maps in TWT
(then in meters) of the Br\'{e}cy depocenter have been computed based
on the interpreted horizons according to the following four steps
methodology, by using the Gverse GeoAtlas module of the Gverse suite
({\textcopyright} Landmark):
\begin{itemize}
\item Computation of grids in TWT for both the TPU and BBU horizons
(these grids were calculated by interpolating the corresponding seismic
horizons and faults).
\item Computation of a thickness map in TWT of the
Carboniferous--Permian sedimentary succession by subtracting the two
previous grids.
\item Calculation of an average interval velocity for the
Carboniferous--Permian deposits from the BTY1, SGS1 and COU1 wells, all
of which have identified Permian deposits. The velocity for each well
is calculated from the thickness of the Permian in meters and in TWT
from the seismic data, giving a value of 3778 m/s. We apply this
``Permian'' velocity to the whole filling, including (i) the
\mbox{intermediate} seismic sequence not fully reached by the three reference
wells, because the seismic facies are similar suggesting similar
lithologies, and (ii) the basal low-frequency/high-amplitude facies,
because (a) they represent much lower thicknesses than the Permian
deposits and are sometimes even missing, and (b) this velocity is very
close to the one estimated in the Arpheuilles well [3615 m/s,
\citealp{Beccalettoetal2015}; location in Figure~\ref{fig1}].
\item Computation of a thickness grid in meters by multiplying the
thickness grid in TWT by the constant velocity of 3778 m/s.
\end{itemize}
As the thickness map is based only on the seismic data (there are no
close outcropping geological contours to be used as hanging points),
the map is hypothetical outside the zones where subsurface data are
present. The resulting map displays the Br\'{e}cy depocenter and its
surroundings with an unprecedented resolution. The maximum thickness of
the Carboniferous--Permian deposits reaches ca. 3900 m along the CH21
seismic line just south of the CH2 line; it is ca. 900 m thicker than
originally expected \citep{Beccalettoetal2015} due to a larger velocity
value and better interpolation controls (because of the availability of
more seismic lines). The minimum thicknesses---about 0 m---are located
northeast of the main depocenter.


\begin{figure*}
\includegraphics{fig06}
\caption{\label{fig6}Thickness map (color range and contour lines) of
the Br\'{e}cy depocenter overlain by the structural map; the geological
background is from the 1:1,000,000 geological map of France 
\citep{BRGM2003}.\break Inset: clockwise rotation of the depocenters.}
\end{figure*}


At the main depocenter, the maximum thicknesses of the \textit{lower},
\textit{intermediate} and \textit{upper seismic sequences} are roughly
900 m, 2500 m, and 500 m respectively. The maximum thicknesses of the
syn-rift deposits (\textit{lower} and \textit{intermediate seismic
sequences,} A, B and C units) and post-rift deposits (\textit{upper
seismic sequence,} D unit) are ca. 3400 m and 500, respectively. In
more detail, the maximum thicknesses of the A, B, C and D units are
approximately 1600~m, 400~m, 500~m and 500~m, respectively.

The main depocenter has an elongated shape trending N60, which is
slightly different to the average N30 trend of the scarce
Carboniferous--Permian faults; in greater detail, it appears that the
successive depocenters represented by the contour lines 3300 m, 3000 m,
2700 m and 2400 m display a\break clockwise rotation of their longest axis
(inset in Figure~\ref{fig6}). This could be the result of a dextral
strike-slip component during the activity of the normal faults and the
opening of the Br\'{e}cy depocenter, either related or not to the
activity of roughly W--E oriented normal faults (not depicted on the
seismic lines as mentioned previously). Broadly speaking, the area
displays a roughly SW--NE and then a W--E ``thick'' axis, located south
of a roughly W--E oriented ``thin'' axis. These features suggest the
combination of two deformational trends---N30 vs. W--E---the
N30 trend is the only one expressed by normal faults. One hypothesis
for this could be that the W--E trend is evidence of similar deeper
trends that have already been identified in the basement
\citep{Baptisteetal2016}.\looseness=1

\subsection{Depositional environment of the four reference wells}

\subsubsection{Electrofacies description}

The lithological evolution can be defined from each studied well using
well-log data and the description of the cuttings.

Based on biostratigraphy data \citep{Bertray11987}, the boundary
between the Carboniferous and Permian is located at 2842~m. The lower
Permian sedimentary succession is characterized by the Autunian facies 
\citep{Bertray11987}, which is composed of, from base to top
(Figure~\ref{fig7}):

\begin{figure*}
\includegraphics{fig07}
{\vspace*{4pt}}
\caption{\label{fig7}Correlations between the BTY1, BCY1 and SGS1 wells
based on stratigraphic cycles.}
{\vspace*{4pt}}
\end{figure*}

\begin{itemize}
\item from 2842 to 2260 m, by conglomerates (only until 2778 m)  and
coarse-grained sandstones characterized by low to medium GR (15 to 80
API) and Sonic (50 to 70 $\rmmu$s/ft) values, thereby
evidencing low-porosity facies, alternating with clayey sandstones with
GR from 80 to 120 API and Sonic below 60 $\rmmu$s/ft, and
brown silty-clay facies characterized by GR from 120 to 200 API and
Sonic from 55 to 110~$\rmmu$s/ft;
\item from 2260 to 1881 m, by brick-red micaceous silty-clay with
traces of anhydrite and dolomite, (GR from 120 to 140 API and Sonic
from 55 to 70 $\rmmu$s/ft), alternating with fine- to very
fine-grained clayey micaceous sandstones (GR from 90 to 120 API and
Sonic from 50 to 60 $\rmmu$s/ft), and white quartzite
sandstones (GR from 45--90 API and Sonic from 40--50
$\rmmu$s/ft).
\end{itemize}
The upper part of the BTY1 well is composed of two distinct facies from
base to top, the name of these formations is found in
\citet{Bertray11987}:
\begin{itemize}
\item the Upper Permian Sandstones Formation, from 1881 m to 1652~m, is
characterized by pink fine- to coarse-grained unconsolidated\break
well-rounded sandstones (GR from 45--75 API and Sonic from 55--65
$\rmmu$s/ft), with few beds of conglomerates described from cuttings
around 1800 m, and alternating with brick-red clay (GR from 90 to 130
API and Sonic from 75--90 $\rmmu$s/ft) or micaceous silty-clay (GR from
75 to 150 API and Sonic from 60 to 75 $\rmmu$s/ft);
\item the Upper Clay Formation, from 1652 to 1314 m, is composed of
brick-red micaceous silty-clay with few beds of green clay in its \mbox{upper}
part (GR from 90 to 150 API and Sonic from 40--80 $\rmmu$s/ft), and
traces of anhydrite; this formation is sandier at the base.
\end{itemize}
The BRC1 well, drilled in \citeyearpar{Brecy11966}, is older than the 
BTY1 well \citeyearpar{Bertray11987} and therefore has only GR in c/s, 
Neutron (c/s) and resistivity well-logs, and thus can only be used to 
provide a qualitative comparison of the GR signal with the other wells 
(Figure~\ref{fig7}). In \citet{Brecy11966},
the Carboniferous is reached below 1802 m, and the sedimentary
succession up to 1707 m has been considered as transitional facies from
the Carboniferous to Permian based on conglomerates interbedded within
brown-red micaceous silty-clay. Based on a previous seismic
interpretation (cf. chapter 4.1) and well-log data, these facies are
considered as Permian. Above 1707 m, BCY1 is considered as
undifferentiated Permian \citet{Brecy11966} and has the following
composition, from base\break to top:
\begin{itemize}
\item from 1707 m to 1565 m: coarse-grained sandstones to conglomerates
interbedded with brown-red micaceous clay with carbonaceous clay around
1676 m, clayey sandstones and dolomitic nodules (highest resistivity
and lower GR values between 1580 and 1651 m);
\item from 1565 m to 1425 m, the sandstones, with sometime dolomitic
cement, are interbedded in micaceous brick-red silty-clay;
\item from 1425 m to 1355 m, the sandstones, interbedded with brick-red
micaceous silty-clay, are pink, sometimes clayey, with a thin
conglomeratic bed at the base of the sandstone level (lowest GR value
and highest resistivity);
\item from 1355 m to 1040 m, the succession is composed of brick-red
micaceous silty-clay, sometimes green and interbedded with a rare and
thin bed of fine-grained clayey sandstones (lowest GR value and highest
resistivity) or traces of anhydrite and gypsum  \citep{Brecy11966}.
\end{itemize}
The SGS1, dated 1964, has only GR (c/s) and resistivity, and is
considered as an undifferentiated Permian succession 
\citep{Georges-sur-Moulon11964} that is characterized from the base
to top by (Figure~\ref{fig7}):
\begin{itemize}
\item from 1639 m to 1537 m, conglomerates and fine- to medium-grained
micaceous white to pink sandstones, sometimes dolomitic, interbedded
with brick-red-brown micaceous and dolomitic silty-clay;
\item from 1537 m to 1316 m, brick-red micaceous silty-clay with
sometimes green clay, rare sandstone levels and traces of\break anhydrite.
\end{itemize}

\subsubsection{Interpretation of depositional environment}

An interpretation of the depositional environment evolution can be
proposed by performing a comparison with adjacent areas. The wells with
thicker and deeper deposits, i.e., BTY1 and BCY1, display different
facies at the base of the succession with coarse-grained and
conglomerate facies interbedded with brown silty clay facies, with
scarce carbonaceous clays. These facies could attest to a sediment
supply attributed to an alluvial fan or delta fan within deeper lake
deposits as described southward in the Lucenay-l\`{e}s-Aix area 
\citep{Ducassouetal2019,Mercuzotetal2021}.

Compared with the core description of COUY1 \citep{Juncaletal2018}, the
mainly red silty-clay facies observed in the upper part of these wells
(unit D, Figure~\ref{fig8}), with some dolomitic beds, can be
considered as a shallow lake. The sandstone and very few conglomerates,
attesting to a sediment supply from a continental area, are attributed
to an alluvial fan sediment supply within the lake
\citep{Juncaletal2018}, the sediment supply is higher in COUY1 than in
the three other wells (Figures~\ref{fig7},~\ref{fig8}). This upper part of the
sedimentary succession shows traces of anhydrite and gypsum from the
description of the cuttings that could reflect an environment with more
evaporation. However, given that only traces were observed and that no
anhydrite has been observed in the shallow lake deposits of the COUY1
well  \citep{Juncaletal2018}, we cannot exclude that these evaporites
originate from the aforementioned Triassic deposits. Furthermore, the
presence of paleosol developments may be indicated by dolomitic nodules
and green clay facies, which could characterize either dolocretes or
hydromorphic paleosols  \citep{Retallack1988,  TaborMontanez2004}, such
as those observed in the COUY1 well \citep{Juncaletal2018}.


\begin{figure*}
\vspace*{-2pt}
\includegraphics{fig08}
\vspace*{-2pt}
\caption{\label{fig8}Correlations between the SGS1, BCY1 and COUY1
wells based on stratigraphic cycles.}
\vspace*{-2pt}
\end{figure*}

The depositional environment of unit B (Figures~\ref{fig7},~\ref{fig8}), compared with
the lithology defined in units A and D, could be considered as fluvial
or an alluvial fan in a lake environment, but with predominantly
coarse-grained sediment from the continental area. The unconsolidated
well-rounded pink sandstones described only at BTY1 (unit C) could be
attributed to aeolian deposits, but due to the lack of core data, this
is difficult to confirm.

\subsection{Sequence stratigraphy correlations}

The correlations have been cross-checked with the seismic lines to
maintain coherency between the well and seismic interpretations. We
propose correlations between the four wells based on the electrofacies
description and the vertical evolution of the \mbox{depositional} environment
(Figures~\ref{fig7},~\ref{fig8}). The most complete Permian sedimentary succession is
observed in the BTY1 well. From 2842~m to around 2215 m, the vertical
evolution from a predominance of conglomerates and sandstones to a
predominance of silty-clay corresponds to unit A on the seismic lines
(Figures~\ref{fig3},~\ref{fig4}) and attests to a general trend towards an open lake
(Figure~\ref{fig7}). The sandstones and conglomerates can be attributed
to alluvial or delta fan deposits in comparison with adjacent areas 
\citep[e.g.,][]{Ducassouetal2019,Mercuzotetal2021,Mercuzotetal2022}.
After this maximum of lake development (Figure~\ref{fig7}), the clay
facies become micaceous, sandier, and display some level dolomites
indicating a shallower lake environment that corresponds to unit B on
the seismic lines (Figures~\ref{fig3},~\ref{fig4}). They vertically evolve to
well-developed sandstone facies reached at 1652~m and correspond to
unit C (Figure~\ref{fig7}). This evolution from the maximum of lake
deposits to the top of the sandstones attests to a general coarsening
upward trend with more sediment supply around 1768 m
(Figure~\ref{fig3}). These sandstones are usually characterized by a
clayey matrix and some conglomerate beds and could be attributed to
fluvial or alluvial fan deposits; their upward evolution until 1652~m
to unconsolidated well-rounded sandstones could suggest aeolian
deposits interbedded within some red-silty clay facies. The general
evolution, which corresponds to the \textit{intermediate seismic
sequence} (Figures~\ref{fig3},~\ref{fig4}), allows to characterize a retrogradation and
progradational trend from unit A to unit B \mbox{(Figure~\ref{fig7}).} Unit C
may characterize the transition to a next retrogradational trend.
However, it is difficult to define without core data, i.e., the
sandstone recorded the transition from a progradational to
retrogradation trend, and particularly if aeolian deposits are
preserved  \citep{Bourquinetal2009,Olivieretal2023}. Moreover, based on
seismic data, a well-defined \mbox{angular} unconformity has been defined at
the top of unit C (i.e., in between the intermediate and upper seismic
sequences; Figures~\ref{fig3},~\ref{fig4}). As a result, unit D is not in continuity with
the previous sedimentary succession. Above 1652~m, the sedimentary
succession shows shallow lake deposits made of clay with some traces of
evaporites, green clay and dolomitic nodules; the latter most probably
represents some paleosol developments. This evolution characterizes a
retrogradational trend (Figures~\ref{fig7},~\ref{fig8}) ended by the Triassic
unconformity and corresponds to the \textit{upper seismic sequence}
(Figures~\ref{fig3},~\ref{fig4}).\looseness=-1

For the BCY1 well, the whole sedimentary succession overlying probable
Carboniferous deposits (top at 1802 m) is thinner than in the BTY1
well. A first trend is observed from conglomerates and sandstones until
a maximum of lake deposits around 1618~m is reached (unit A; 
Figures~\ref{fig7},~\ref{fig8}), correlated with the basal retrogradational trend of BTY1. A
progradational trend, with well-developed sandstones and conglomerates
evolving to fine-grained micaceous sandstones is observed until 1425 m
(unit B; Figures~\ref{fig7},~\ref{fig8}). Unit C is less developed in this borehole.
Above the angular unconformity (1385~m), the same retrogradational
trend as in BTY1 is observed, but with more sandstone and conglomerate
deposits attesting to an area with more sediment supply (unit D;
Figures~\ref{fig7},~\ref{fig8}).

In the SGS1 and COU1 wells, the Permian sedimentary succession has a
reduced thickness corresponding to a general retrogradational trend
from conglomerates and sandstones to silty-clay facies. The latter
corresponds to the same facies belonging to the upper retrogradational
trend of the three other wells (unit D; Figures~\ref{fig7},~\ref{fig8}), with the
occurrence\break of more sandy beds in COU1 (Figure~\ref{fig8}). In this
well, eight retrogradational--progradational cycles were described based
on core analyses within this retrogradational trend
\citep{Juncaletal2018}, but in the absence of core data it is
impossible to propose correlations at this high-resolution scale with
the BTY1, BCY1 and SGS1 wells.

\section{Discussion}

\subsection{Tectonic evolution of the Br\'{e}cy depocenter---regional
comparisons}\label{sec5.1}

The new seismic reprocessing greatly improves the understanding of the
structural style of the Br\'{e}cy depocenter, leading us to discuss the
modality of its syn- and post-rift tectonic evolution
(Figure~\ref{fig9}), and to look for similar patterns in other LOCPB.


\begin{figure*}
\vspace*{3pt}
\includegraphics{fig09}
\caption{\label{fig9}Present-day situation and tectonic evolution of
the Br\'{e}cy depocenter, using the representative composite CH41-20-2
seismic line. The syn-rift, uplift and post-rifts phases are
successively depicted; the ages come from comparison with the
Lorraine-Saar-Nahe Basin; see text for explanations.}
\vspace*{4pt}
\end{figure*}

\subsubsection{Syn-rift stage}
The new interpretation confirms and refines
the multi-step widening scenario for the Br\'{e}cy depocenter,
controlled by the westward migration of a few successive active normal
faults and their related depocenters, thinning and onlapping each other
eastward (Figure~\ref{fig9}). Some of the normal faults may be
connected to deep detachments located in the pre-Stephanian basement,
giving them a listric geometry at the time of the deposition
(Figure~\ref{fig9}). In fact, upper crustal extensional detachments in
the footwall of LOCPB are a common structural feature around the Massif
Central 
\citep{EchtlerMalavieille1990,Burgetal1994,Faure1995,Gardienetal2022},
and such detachments have been described close to the study area in the
Autun and La Serre regions 150 km east of the Br\'{e}cy depocenter
\citep{Chouletetal2012}.\looseness=1

This induced the creation of an accommodation space that recorded a
retrogradational trend characterized by well extended lacustrine
deposits (unit A), followed by a progradational trend (unit B), due to
increasing sediment supply in this subsiding context, before a probably
new retrogradational trend (unit C) was abruptly ended by the angular
unconformity (Figures~\ref{fig7},~\ref{fig8}). 

The intersecting N20-trending CH3 and CH21 lines are parallel to the
main structural trend (Figure~\ref{fig6}). There, the eastward facing
normal faults are sub-parallel to the lines, so that they appear as
flat lens-shape features subparallel to the BBU, disturbing the
geometry of its reflectors. Deeply seated highly energetic reflectors
in the basement below the BBU also occur and are similarly interpreted
as potential extensional detachments planes.

All the above observations point to a syn-tectonic filling of the basin
in an overall extensive tectonic regime during the latest
Carboniferous--early \mbox{Permian,} characterizing the syn-rift stage of the
Br\'{e}cy depocenter. The maximum thickness of the syn-rift stage is
ca. 3400 m.

\subsubsection{Post-rift stage}
The syn-rift activity is followed by a
post-rift stage, whose occurrence is documented by the toplap geometry
of the horizons of the syn-rift stage (mainly belonging to units B and
C) below the first deposits of the overlying unit D (\textit{upper
seismic sequence}, that corresponds to a new stratigraphic cycle where
only its retrogradational phase is recorded due to an unconformity
before Triassic sediment preservation) (Figures~\ref{fig5},~\ref{fig6},~\ref{fig9}). The toplap
geometries, observed in the whole area, provide evidence for the uplift
and subsequent erosion of the syn-rift sediments before the deposition
of the post-rift stage s.s. The latter is roughly isopach throughout
the Br\'{e}cy depocenter (ca. 0.25 s TWT, corresponding to ca. 500~m),
and extends \mbox{laterally} beyond the syn-rift stage. Another typical
feature characterizing the post-rift stage is the absence of faults
controlling the sedimentation. When considered together, these features
suggest that the post-rift stage developed under a regional subsidence
regime, which is certainly related to the late orogenic thermal
relaxation of the hot Variscan crust  
\citep[e.g.,][and references
therein]{AverbuchPiromallo2012,Vanderhaegheetal2020}. 
Toplaps of horizons belonging to the post-rift stage below
the Triassic deposits suggest that the whole Br\'{e}cy depocenter
(i.e., syn-rift and post-rift stage) was then slightly tilted and
subsequently eroded before the arrival of the first Triassic sediments.
This erosional event may explain the relative low thickness of the
post-rift stage (well depicted in the CH21, CH3 and CH41-20-2 seismic
lines, Figures~\ref{fig5},~\ref{fig6}). \citet{Beccalettoetal2015} checked and found
similar geometrical and structural patterns in the adjacent Contres and
Arpheuilles depocenters (thick syn-rift record beneath a thin isopach
post-rift\break record).

The transition from a tectonic subsidence regime accommodated by
discrete faults (syn-rift stage) to a regional subsidence regime
(post-rift stage) through an uplift phase has not been documented yet
in other French LOCPB, except in the Lorraine region (location in
Figure~\ref{fig1}). There, the latest Carboniferous--lower Permian
syn-rift deposits are also controlled by listric normal faults rooting
downward in deep decollement levels  \citep{Hemelsdaeletal2023}. They
are then eroded following their uplift and tilting, before the arrival
of the early to middle Permian post-rift sediments  \citep{Henk1993, 
Stollhofen1998}. The latter are then themselves eroded (toplap
geometries below the Triassic deposits) before the deposition of the
first Triassic sediments, as in the Br\'{e}cy area 
\citep{Hemelsdaeletal2023}. Further east in Germany, the Upper
Rotliegend units of the Saar-Nahe (eastward prolongation of the
Lorraine Basin), Thuringian Forest and Saale basins may also record
these post-rift events  \citep{HertleLittke2000,Schafer2011,SchneiderRomer2010}. In the literature, the tectonic phase
responsible for the uplift is attributed to the end of early Permian
Saalian phase  
\citep[e.g.,][]{McCannetal2008a,Pharaohetal2010,Stille1924,Ziegler1990},
originally described in Central Germany, where folding and local
tectonic inversion occur  \citep{HertleLittke2000,Kneuper1976}. In
addition to the observations made in the Br\'{e}cy depocenter, the
shortening phases recognized in some French LOCPB may also be related
to this poorly known Saalian phase 
\citep{Blesetal1989,BonijolyCastaing1983, 
Gelardetal1986,GennaDebriette1994,MattauerMatte1998}. Going back to the
Br\'{e}cy area and according to this scenario, the base of the
post-rift stage (unit D) would correspond to the superposition of two
regional unconformities, i.e., the post-rift unconformity and the
post-Saalian\break unconformity.

Lastly, it is unsurprising to not find explicit evidence of such a
post-rift event in the French outcropping LOCPB, as the related
deposits have certainly been partly removed by subsequent Meso-Cenozoic
erosion  \citep{Barbarandetal2013,Guillocheauetal2000}. They are best
fully described in subsurface settings beneath their Mesozoic cover,
where the Permian deposits (including post-rift sediments) are still
preserved and therefore accessible to seismic observations; in other
words, the subsurface prolongation of other LOCPB around the Massif
Central  (e.g., Brive, Rodez, Saint-Affrique) may indeed preserve such
post-rift deposits, even though they have not yet been identified. The
recognition of a syn-rift stage followed by a regional uplift and a
post-rift stage in the late Carboniferous--Permian Br\'{e}cy depocenter
is of great importance, as it allows to make a geodynamic connection
with known LOCBP in eastern France and Germany. It confirms that the
development of late Variscan basins is controlled by a large-scale
(western Europe) underlying geodynamic event  [e.g., the lithospheric
delamination and removal of the Variscan mantle roots, e.g.,
\citealp{Arnoldetal2001,AverbuchPiromallo2012}]. Later, the arrival of
the first Triassic sediments attests to the subsequent long-term
Mesozoic thermal subsidence of the lithosphere, which led to the
formation of the Paris Basin as a post-Variscan sag basin 
\citep{AverbuchPiromallo2012,Prijacetal2000,Robinetal2000}.
\looseness=1

\subsection{The Br\'{e}cy depocenter was belonging to a much larger basin---local vs. regional correlations}

In addition to structural and tectonic aspects, which are difficult to
understand without seismic data, another way to look for lateral
correlations of the Br\'{e}cy depocenter with neighboring basins is to
track the thicknesses of preserved end-Carboniferous--Permian deposits,
their facies and paleoenvironments, and age similarities. In this
section, we aim to connect the Br\'{e}cy depocenter to the LOCPB of the
northern Massif Central, before looking for wider-scale regional
correlations.

First, as already demonstrated, the Br\'{e}cy depocenter is undoubtedly
connected westward to the Contres depocenter (which itself is connected
to the Arpheuilles depocenter), as shown by the continuity of the
Permian reflectors on the seismic lines between the three depositional
areas \citep{Beccalettoetal2015}. It is difficult to assess the
occurrence of \mbox{end-Carboniferous--Permian} deposits north to northeast of
the Br\'{e}cy depocenter as there are no available seismic lines; this
is also the case for its southwest prolongation. The few available
spatially dispersed wells in these areas indicate that the Triassic
deposits rest directly on basement rocks without intervening
Carboniferous--Permian deposits (weathered granite in the Pougues-1 and
Sainte-Colombres-1 wells, micaschists in the M\'{e}netou-1 wells;
location in Figure~\ref{fig2}). It is however impossible to
definitively conclude on the complete absence of Carboniferous--Permian
rocks to the north of the Br\'{e}cy depocenter based on such scarce
data, as they may occur in between the wells. It could for instance be
the case along the roughly N--S Loire-Sancerre fault zone and its
northward prolongation as suggested by \citet{Delmasetal2002}
(Figure~\ref{fig10}). In addition, recent work done by the 45-8 Energy
Company and BRGM leads to the recognition of new small subsurface
Permian deposits southwest of the La Machine area 
\citep{Hauvilleetal2021, Jacobetal2021}.
This very local Permian occurrence may be connected to the Br\'{e}cy
depocenter, but their connection is not yet known.\looseness=-1

\begin{figure*}
\vspace*{-2pt}
\includegraphics{fig10}
\vspace*{-1pt}
\caption{\label{fig10}Estimated extent of the ``northern Massif Central
Carboniferous--Permian Basin'' on a present-day geological map; the
geological background is from the 1:1,000,000 geological map of France
\citep{BRGM2003}.}
\vspace*{-2pt}
\end{figure*}

Further south, it is strongly tempting to extend the Br\'{e}cy
depocenter to the outcropping Aumance Basin, although there is no
direct evidence for this (no seismic line or well data). Extensive
research has been carried out in the Aumance area for the purpose of
uranium exploration, resulting in detailed sedimentological and
structural descriptions based on numerous borehole and scarce seismic
data \citep{MathisBrulhet1990}. There, the Buxi\`{e}res Formation with
gray Autunian facies has been dated close to the Carboniferous--Permian
transition [298.59 ${\pm}$ 0.35~Ma, CA-ID-TIMS method,
\citealp{Mercuzotetal2023}; Figure~\ref{fig11}]. This formation is
overlain by red Autunian alluvial to lacustrine facies (Reni\`{e}re
Fm.). Both facies are similar to those belonging to the intermediate
seismic sequences (units A and B) of the Br\'{e}cy depocenter. When
considered together, these facies provide evidence for a syn-rift stage
\citep{MathisBrulhet1990,  Paquette1980}. Note that the Aumance is not
located along strike of the N30-trending Br\'{e}cy depocenter, but
instead is shifted to the east, such that the former should not be seen
as the southward prolongation of the latter but as a parallel basin
(Figure~\ref{fig10}).\looseness=-1

Eastward, recent works from \citet{Mercuzotetal2021,Mercuzotetal2022}
suggest the connection of the Decize-La-Machine (including the
Lucenay-l\`{e}s-Aix depocenter), Aumance, Autun and Blanzy-Le Creusot
Stephano-Autunian basins: their lacustrine environments (from alluvial
fan/fan delta to deep lake) and depositional ages are similar [latest
Carboniferous to early Permian;
\citealp{Ducassouetal2019,Pellenardetal2017}]. Here again, the Permian
facies are similar to the intermediate (units A) seismic sequences of
the Br\'{e}cy depocenter, and they display seven
retrogradational--progradational cycles within a general
retrogradational trend ended by an unconformity overlain by Triassic
deposits \citep{Mercuzotetal2021}. Even though we observed the same
general retrogradational evolution in the Br\'{e}cy area (unit~A), it
is not possible to make a detailed comparison between these two areas
because no age is available at the top of the Permian in the
Decize-La-Machine area. Consequently, the absence of sandstones in
units B and C (i.e., mainly a progradational trend) in
Decize-La-Machine is either due to the erosion of these units or to a
local sediment supply in the Br\'{e}cy area, which is not recorded in
Decize-La-Machine. However, a significant difference lies in the
cumulative thickness of the deposits, which are unsurprisingly much
lower in the outcropping basins (ca.~1 to 1.2 km) compared to the
Br\'{e}cy depocenter (2.5~km; Figure~\ref{fig11}). Such large
cumulative thicknesses are only reported from the Blanzy-Le Creusot
composite basin, where more than two kilometers of Stephano-Autunian
deposits are described from coal exploration boreholes
[Figure~\ref{fig11}, \citealp{BRGM1989,Gand2003}]. Further east, structural
connections between the Autun, La Serre, Burgundy and southern Vosges
basins have also been proposed \citep{Chouletetal2012}.\looseness=1
\pagebreak

\onecolumngrid
\begin{sidewaysfigure*}[p!]
{\vspace*{20pc}}
\includegraphics{fig11}
\caption{\label{fig11}Compilation of thicknesses and ages of the main
LOCPB in France and Germany, including the Br\'{e}cy depocenter as a
part of the ``northern Massif Central Carboniferous--Permian Basin''.}
\end{sidewaysfigure*}
\twocolumngrid

All these observations and interpretations, correlating neighboring
basins to each other, imply that these present-day isolated outcropping
and subsurface basins, represented by the Stephano-Autunian facies,
were part of a larger basin during the latest Carboniferous--early
Permian. This larger basin encompassed the subsurface Arpheuilles,
Contres and Br\'{e}cy depocenters and the eastward smaller outcropping
basins, spanning more than four hundred kilometers in length, with a
roughly W--E trend (``northern Massif Central Carboniferous--Permian
Basin'', Figure~\ref{fig10}). These regional connections are also
confirmed by paleobiogeography data from freshwater shark remains found
in the Stephano-Autunian sediments, suggesting connections between all
the French LOCPB of the northern Massif Central [including the Aumance,
Autun and Blanzy-Le Creusot basins as well as on a much larger scale as
far away as Germany;
\citealp{Luccisanoetal2021,Fischeretal2013,Schneideretal2020}].

It is important to note that the area depicted in Figure~\ref{fig10}
(pale blue color) may represent a \textit{minimum} extent of this wide
basin. Indeed, some upper Carboniferous and lower Permian deposits have
certainly been removed through time and are therefore lacking, due to
(i) the erosion phases known before the arrival of the Triassic
sediments [as discussed above and in \citealp{Beccalettoetal2015}], and
(ii)~the Meso-Cenozoic uplift of the Massif Central area and subsequent
erosion  \citep{Barbarandetal2013,Guillocheauetal2000}.

On a wider scale, unlike the northern Massif Central, several-km-thick
Stephano-Autunian deposits are encountered in several LOCPB of the
southern Massif Central. These basins, such as Brive,
Decazeville-Rodez, Carmaux-La Gr\'{e}sine, Saint-Affrique or
Graissessac-Lod\`{e}ve, display similar Stephano-Autunian facies and
ages to those observed in the Br\'{e}cy depocenter and connected
smaller basins, and reach thicknesses of up to four kilometers
[Figure~\ref{fig11}; 
\citealp{BRGM1989,PochatVanDenDriessche2011,Poujoletal2023}]. Due to the lack of
subsurface seismic data, their extensional structural control is rarely
imaged  [\citealp{Serranoetal2006} for a counterexample in the
Aquitaine Basin] and is inferred from indirect observations
\citep{Chenetal2006,  PochatVanDenDriessche2011}.

Such thick series are also known in the LOCPB of Germany, for example
in the Saar-Nahe and Thuringian Forest basins 
\citep{SchneiderRomer2010,Schneideretal2020}. There, studies benefit
both from favorable outcrop conditions and a large range of subsurface
dataset including seismic lines. Consequently, the sedimentary facies
and \mbox{paleoenvironments} are well constrained, the lithostratigraphic and
extensional structural features are well described, and recent absolute
depositional ages are available  [\citealp{Henk1993,Stollhofen1998, 
Schafer2011,  Lutzneretal2020,Voigtetal2022}; Figure~\ref{fig11}].
Based on all of these characteristics, they can be considered as other
robust reference basins for large-scale LOCPB correlations.

All other LOPCB occurrences in western Europe generally occur in
deformed areas as small, isolated patches with low preserved
thicknesses, like in the Pyrenees (French and Spanish sides),
Cantabria, Sardinia and the Alps and surrounding areas (France, Italy,
Switzerland) 
\citep{BarnolasChiron1996,BRGM1989,Capuzzoetal2003,Cassinisetal2012, 
KnightAlvarez-Vazquez2021,Lloretetal2018,Pittauetal2002,  Toutin1980}.
We should also mention the supposedly wider basins with uncertain
boundaries, thickness and age attributions that are also known offshore
based on seismic data  [e.g., Bay of Biscay, Western Approaches;
\citealp{Boisetal1991,  Ziegler1990}]. These characteristics make all
these basins unsuitable for reliable wide-scale comparisons, and cannot
be considered as reference basins.

Lastly, arguments in favor of the Br\'{e}cy depocenter being part of a
new wider-scale reference basin (which would be useful for further
large-scale correlations, and located in between the two poles
represented by the LOCPB of the southern Massif Central and Germany)
include: its large preserved thickness (up to 3.9 kilometers),
structural record (syn- and then post-rift stages) and age calibration
(paleontological and absolute ages from deep wells and comparison with
close connected basins with absolute temporal calibrations).

\section{Summary and conclusion}

The structural pattern, thickness, tectonic evolution and depositional
environments of the late Carboniferous--Permian Br\'{e}cy depocenter are
revealed based on the reprocessing and interpretation of 115 km of
vintage seismic lines combined with deep well data.

The present-day geometry of the Br\'{e}cy depocenter is controlled by
seven major and eleven minor eastward dipping normal faults; all these
faults were active during the Meso-Cenozoic period, and some of them
were possibly connected to deep detachments active during the late
Carboniferous--Permian \mbox{history.} The maximum thickness of the Br\'{e}cy
depocenter is reevaluated and estimated to be 3900 m.

The filling indicates a thick late Carboniferous--early Permian syn-rift
stage overlain by a thin early to middle post-rift stage, similar to
the tectonic evolution of the northeastward basins in Lorraine and
Germany. The facies generally characterize lake environments, with
occurrences of fluvial, alluvial and delta fan deposits. They mainly
display a retrogradational--progradational pattern during the syn-rift
stage, and a retrogradational pattern during the post-rift stage. The
Br\'{e}cy area is therefore the thickest depocenter known in the
northern Massif Central by far. We propose that it was part of a larger
basin during the latest Carboniferous--early Permian times, with a
roughly W--E trend spanning more than four hundred kilometers in length
in the northern Massif Central region. Strictly speaking, all the other
so-called outcropping basins (Aumance, Decize-La-Machine, etc.) should
be considered as depocenters, as they belong to a wider scale basin.

Because of the above features, this ``northern Massif Central
Carboniferous--Permian Basin'' may be seen as a reference LOCPB,
comparable to the basins known in the southern Massif Central, eastern
France and Germany. Lastly, our work emphasizes the need for a further
multi-method approach to explore LOCBP, especially their subsurface
occurrences, as it is crucial to have a full view of these basins that
contain information that is inaccessible to their outcropping remnants.
It definitely highlights that LOCPB were obviously much wider and
thicker than previously expected, with potential effective impacts on
the understanding of the late Variscan tectonic and climatic evolution
in France and western Europe.\looseness=-1

\section*{Declaration of interests}
The authors do not work for, advise, own shares in, or receive funds
from any organization that could benefit from this article, and have
declared no affiliations other than their research organizations.

\section*{Acknowledgements}

This publication was initiated within the framework of the ICDP
Deepdust project and was supported by the BRGM's Research Division. The
authors gratefully acknowledge Fran\c{c}ois Baudin, coordinator of this
special issue dedicated to Jean Dercourt, for his patience and
perseverance. They thank the two anonymous reviewers for their
constructive comments, which have greatly improved the manuscript. The
authors also thank Sara Mullin for proofreading the English content.

\nocite{BreitkreuzEhling2009,
BRGM1970,
BRGM1976,
Bruguieretal2003,
Durand2014,
Gandetal1996,
Goloubinoff1979,
Hubneretal2011,
Roigetal2001,
Rolando1988,
Schneideretal2006}

\back{}

\bibliographystyle{crgeos}
\bibliography{crgeos20230749}
\refinput{crgeos20230749-reference.tex}

\end{document}