Figure - available from: Nature Communications
This content is subject to copyright. Terms and conditions apply.
Garrison Basin multicore reconstructions and corresponding stacked records. Individual core Mg/Ca (mmol/mol) and δ¹⁸Oc data (‰, VPDB), and δ¹⁸Osw (‰, VSMOW) and SST (°C) reconstructions (blue–MCA, red- MCB, yellow–MCC) plotted with median and 68% uncertainty envelope incorporating age, analytical, calibration, and sampling errors (a-d) along with corresponding median stacked records with 68% and 95% confidence bounds (e-h). Diamonds in a and e indicate stratigraphic points sampled for radiocarbon. Gray histogram in g is the probability distribution for a changepoint in the δ¹⁸Osw time series. Orange circle in g is the mean of available δ¹⁸Osw measurements in the GOM and orange line in h is observed monthly mean SST with uncertainty envelope calculated using a Monte Carlo procedure that simulates foraminiferal sampling protocol. Purple line in h is the 100-year running correlation between SST and δ¹⁸Osw with corresponding uncertainty with shaded boxes indicating correlations with r > 0.7 (p < 0.001), which is the basis for identifying time periods where Loop Current and associated processes are relevant
Source publication
Surface-ocean circulation in the northern Atlantic Ocean influences Northern Hemisphere climate. Century-scale circulation variability in the Atlantic Ocean, however, is poorly constrained due to insufficiently-resolved paleoceanographic records. Here we present a replicated reconstruction of sea-surface temperature and salinity from a site sensiti...
Similar publications
High-resolution seawater δ18O records, derived from coupled Mg/Ca and benthic δ18O analyses, can be used to evaluate how global ice volume changed during the mid-Pleistocene transition (MPT, ca. 1250–600 ka). However, such seawater δ18O records are also influenced by regional hydrographic signals (i.e., salinity) and changes in deep-ocean circulati...
Citations
... Northward shifts in the loop current lead to increased tropical cyclone precipitation for the Northern GOM (Bregy et al., 2022). Inferred northward migration of the loop current, documented by Thirumalai et al., 2018 andBregy et al., 2022 over the last ~300 years could be a suitable explanation for the coupling observed between central TX intense TC landfalls and extreme, positive PDSI events. ...
... The AMOC, a crucial regulator of the climate system, also displays variability on this multicentennial timescale (Kissel et al. 2013;Thornalley et al. 2013). Consequently, MCV of the AMOC has been proposed as a potential driver of the climate system's MCV (McDermott et al. 2001;Oppo et al. 2003;Hall et al. 2004;Miettinen et al. 2012;Chabaud et al. 2014;Ayache et al. 2018;Thirumalai et al. 2018). ...
Using a CESM1 control simulation, we conduct a follow-up study to advance our earlier theoretical research on the multicentennial oscillation (MCO) of the Atlantic meridional overturning circulation (AMOC). The modeled AMOC MCO primarily arises from internal oceanic processes in the North Atlantic, potentially representing a North Atlantic Ocean-originated mode of AMOC multicentennial variability (MCV) in reality. Specifically, this AMOC MCO is mainly driven by salinity variation in the subpolar upper North Atlantic, which dominates local density variation. Salinity anomaly in the subpolar upper ocean is enhanced by the well-known positive salinity advection feedback that is realized through anomalous advection in the subtropical-subpolar upper ocean. Meanwhile, mean advection moves salinity anomaly in the subtropical intermediate ocean northward, weakening the subpolar upper salinity anomaly and leading to its phase change. The salinity anomalies have a clear three-dimensional life cycle around the North Atlantic. The mechanism and timescale of the modeled AMOC MCO are consistent with our earlier theoretical studies. In the theoretical model, artificially deactivating either the anomalous or mean advection in the AMOC upper branch prevents it from exhibiting AMOC MCO, underscoring the indispensability of both the anomalous and mean advections in this North Atlantic Ocean-originated AMOC MCO. In our coupled model simulation, the South Atlantic and Southern Ocean do not exhibit variabilities synchronous with the AMOC MCO; the Arctic Ocean’s contribution to the subpolar upper salinity anomaly is much weaker than the North Atlantic. Hence, this North Atlantic Ocean-originated AMOC MCO is distinct from the previously proposed Southern Ocean-originated and Arctic Ocean-originated AMOC MCOs.
... warming rate of 0.01°C y 1 ± <0.01°C y 1 from 1830 to 2011 C.E., and the multispecies composite shows a long-term warming trend of 0.002°C y 1 ± 0.0004°C y 1 from 1733 to 2011 C.E. This is consistent with rates of warming reported from other marine proxy records of SST from the Gulf of Mexico-Caribbean region, which indicate that LIA SST was between 1°C and 3°C cooler than modern (DeLong et al., 2014;Kilbourne et al., 2008;Lund & Curry, 2006;Richey et al., 2007Richey et al., , 2009Thirumalai et al., 2018). The warming interval beginning at 1830 C.E. in the Orbicella Sr/Ca-SST composite record is also consistent with, but more rapid than the warming trend derived from the annually resolved SST proxies spanning the Caribbean and sub-tropical Atlantic (i.e., 0.0006°C y 1 ± 0.0001°C y 1 , Tierney et al., 2015). ...
Long‐lived Atlantic coral species like Orbicella faveolata are important archives of oceanographic change in shallow, marine environments like the Florida Keys. Not only can coral‐based records extend for multiple centuries beyond the limits of the instrumental record, but they can also provide a more accurate representation of in situ conditions than gridded interpolated sea‐surface temperature (SST) products for nearshore reef environments. We use the coral Sr/Ca paleothermometer to produce a 150‐year (1830–1980 C.E.) monthly SST reconstruction from an O. faveolata colony collected in the Marquesas Keys, FL, USA. An important feature of our record is a significant 20th‐century warming trend in winter SSTs. We hypothesize that the winter warming trend was driven partially by a decrease in upwelling associated cyclonic eddies spinning off the Florida Current. A long‐term weakening of winter Florida Current transport over the 20th century could be responsible for decreased cyclonic eddy formation in the Florida Straits. Another feature of the record is pronounced multidecadal fluctuations of mean annual warming and cooling in the record, which correspond to Atlantic Multidecadal Variability (AMV), with the AMV lagging behind western Florida Keys temperatures by 5–11 years. Strong coherence between coral‐based SST reconstructions in the western Florida Keys with broader scale Atlantic oceanographic trends over the past century suggests a common driver of regional SST variability.
... Numerous studies have used Mg/Ca ratio in foraminifera tests to reconstruct SST in the tropics at different time scales and localities (Nürnberg et al., 2000;Lea et al., 2000;Dekens et al., 2002;Anand et al., 2003;Gray et al., 2018;Thirumalai et al., 2018;Richey et al., 2019;Saenger and Evans, 2019). In particular, the development of SST reconstructions at a glacial-interglacial scale in the Gulf of Mexico might provide information on the temporal dynamics of the Loop Current, its relationship to the variability of the Mississippi River discharge, and the evolution of the Western Hemisphere warm pool (WHWP) . ...
... The development of the WHWP partially depends on the Intertropical Convergence Zone (ITCZ) migration and the climatic rhythms at the orbital to suborbital scales between the Atlantic Ocean, Caribbean Sea, and Gulf of Mexico (Wang and Enfield, 2001). There are paleo-SST records from Orca Basin (Flower et al., 2004;Poore et al., 2004), DeSoto Canyon Nürnberg et al., 2008), the Florida Straits (Bahr et al., 2011;Schmidt et al., 2012), Pigmy Basin (Richey et al., 2007;Richey et al., 2012), Bryant Canyon (Antonarakou et al., 2015), the Garrison Basin (Thirumalai et al., 2018;Thirumalai et al., 2021;Jentzen et al., 2018), and the western boundary in the North Atlantic (Bahr et al., 2013). As for the time covered, most studies are focused on the last 100,000 years and are very scarce in previous Pleistocene glacial-interglacial periods. ...
Analyses of Mg/Ca in planktonic foraminifera Globigerinoides ruber (white) tests were made to evaluate burial influences and preservation before reconstructing paleotemperatures. We studied samples from the eastern Gulf of Mexico marine sediment core EN32-18PC at a glacial-interglacial scale. The core chronology is based on planktonic foraminifera biostratigraphy and δ18OG. ruber stratigraphy, showing that the core spans the biozones V2 - Y and the Marine Isotope Stages (MIS) 9 to 4 (i.e., 319.3 to 66.0 ka). Based on Scanning Electron Microscopy (SEM) and LA-ICP-MS analysis of Mg/Ca, Sr/Ca, Mn/Ca, Fe/Ca, and Mn/Fe ratios, we found evidence of test alteration by early diagenesis. Some tests show material infillings, pore reduction, polygonal patterns between pores, and wall thickening by calcareous overgrowth. Values of Mg/Ca >6 mmol/mol and δ18O > 2‰ must be carefully evaluated before reconstructions. Values of Mn/Ca >0.6 mmol/mol and Fe/Ca >0.2 mmol/mol indicate the adherence of clays or coatings (Mn or Fe oxides). Nevertheless, we did not find evidence of severe diagenesis in the tests (like recrystallization or euhedral crystals grown by inorganic precipitates); thus, their alteration degree is incipient, and calcite overgrowth is the most severe diagenetic effect during MIS-6 and early MIS-5. After calculating paleotemperatures, we obtained SST values comparable to other cores in the region (i.e., 24 °C - 31 °C). The paleotemperature record fluctuates over MIS 9 to 5, in agreement with the insolation intensity cycle controlled by precession. However, suborbital trends could not be reconstructed due to the studied effects imprinted on the specimens.
... The question about the real existence of this centennial variability in the Atlantic is all the more important that in our pseudo-proxies experiment, it corresponds to a significant part of the SIR-LIMs scores in the Atlantic. Unfortunately, the instrumental period appears to be too short to easily dissociate this range of low-frequency variability from the external forcing trends during the instrumental period but stay potentially real since the JEBRI AND KHODRI 10.1029/2022MS003094 20 of 26 proxies analysis tends to suggest the existence of this kind of Atlantic centennial variability in the past (Gray et al., 2004;Thirumalai et al., 2018). Finally, our results also suggest that the SIR-LIM PDA can be considered a tool to allow the production of low-cost reconstructions following dynamical constraints deduce from different GCMs that will help shed some light on the discrepancies found among currently available reconstructions for the last 2000 years (PAGES2k Consortium, 2019). ...
Proxy records that document the last 2000 years of climate provide evidence for the wide range of the natural climate variability from inter‐annual to secular timescales not captured by the short window of recent direct observations. Assessing climate models ability to reproduce such natural variations is crucial to understand climate sensitivity and impacts of future climate change. Paleoclimate data assimilation (PDA) offers a powerful way to extend the short instrumental period by optimally combining the physics described by General Circulation Climate Models (GCMs) with information from available proxy records while taking into account their uncertainties. Here we present a new PDA approach based on a sequential importance resampling (SIR) Particle filter (PF) that uses Linear Inverse Modeling (LIM) as an emulator of several CMIP‐class GCMs. We examine in a perfect‐model framework the skill of the various LIMs to forecast the dynamics of the surface temperatures and provide spatial field reconstructions over the last millennium in a SIR PF. Our results show that the LIMs allow for skillful ensemble forecasts at 1‐year lead‐time based on GCMs dynamical knowledge with best prediction in the tropics and the North Atlantic. The PDA further provides a set of physically consistent spatial fields allowing robust uncertainty quantification related to climate models biases and proxy spatial sampling. Our results indicate that the LIM yields dynamical memory improving climate variability reconstructions and support the use of the LIM as a GCM‐emulator in real reconstruction to propagate large ensembles of particles at low cost in SIR PF.
... Although millennial-scale climate change in the North Atlantic region during the late Holocene was dominated by externally forced cooling (Wanner et al., 2008;Marcott et al., 2013), paleoclimate records from the subtropical western Atlantic indicate higher frequency sea-surface temperature (SST) variability over shorter centennial timescales. In particular, SST reconstructions from the northern Gulf of Mexico and Florida Straits reveal a period of sustained SST warming between~3500 and 2000 years B.P. (Figures 3A, B; Poore et al., 2004;Lund and Curry, 2004;Thirumalai et al., 2018;Thirumalai et al., 2021). Both sites are located at the northern extent of the Atlantic Warm Pool (AWP) (Wang and Enfield, 2001), a feature that forms annually when warm Caribbean water masses are transported around the Florida Peninsula and into the subtropical western Atlantic via the Loop and Florida currents (Lee et al., 1995;Lund and Curry, 2004; Figure 1A). ...
... longer present on the NRC off northern Broward County with the exception, perhaps, during a brief period at 975 years B.P. (indicated by a single A. palmata age). The regional extirpation of these corals corresponds to a long-term shift towards cooler SSTs throughout the North Atlantic (Wanner et al., 2008), including at the northern AWP sites in the Gulf of Mexico and Florida Straits (Figures 3A, B;Poore et al., 2004;Lund and Curry, 2004;Thirumalai et al., 2018;Thirumalai et al., 2021). Significantly cooler temperatures relative to 3500-2000 years B.P. may have diminished the warm water buffering effect of the Florida Current, thus limiting the ability of A. palmata and Orbicella spp. ...
... Not surprisingly, A. palmata and Orbicella spp. corals were absent on the NRC during the Little Ice Age (1450 to 1850 CE) when SSTs in the northern AWP region were punctuated by cooling as much as 2°C compared to instrumental records spanning between 1970 and 2010 CE (Richey et al., 2007;Thirumalai et al., 2018). ...
The Holocene reefs off southeast Florida provide unique insights into the biogeographical and ecological response of western Atlantic coral reefs to past climate change that can be used to evaluate future climate impacts. However, previous studies have focused on millennial-scale change during the stable mid-Holocene, making it difficult to make inferences about the impact of shorter-term variability that is relevant to modern climate warming. Using uranium-series dating of newly discovered subfossil coral rubble deposits, we establish a new high-resolution record of coral community development off southeast Florida during a period of variable climate in the late Holocene. Our results indicate that coral communities dominated by reef-building Acropora palmata and Orbicella spp. persisted in the nearshore environments off southeast Florida ~75 km north of their primary historical ranges between ~3500 and 1800 years before present. This timing coincides with regional warming at the northern extent of the Atlantic Warm Pool, suggesting a likely link between regional oceanographic climate and the expansion of cold-sensitive reef-building coral communities to the high-latitude reefs off southeast Florida. These findings not only extend the record of coral-reef development in southeast Florida into the late Holocene, but they also have important implications for future range expansions of reef-building coral communities in response to modern climate change.
... In contrast, other researchers have concluded that the AMOC was not significantly altered during the LIA. They conclude that the observed hydroclimatic impacts were instead related to changes in wind regime (e.g., Moreno-Chamarro et al., 2015;Moreno-Chamarro, Ortega, et al., 2017;Moreno-Chamarro, Zanchettin, Lohmann, Luterbacher, et al., 2017;Thirumalai et al., 2018). ...
... It should be noted though that determining the underlying causes for phases of the LIA is an area of active investigation with other research suggesting that an increase in TSI may not have significantly influenced the LIA (e.g., Chiodo et al., 2019;Neukom et al., 2019;Schurer et al., 2013). Further, there is contradictory research that suggests that the AMOC (and deepwater formation) may not necessarily have been altered during the LIA, but rather that the observed hydroclimate impacts occurred due to Atlantic-wide surface-circulation changes related to changes in the wind regime (e.g., Moreno-Chamarro et al., 2015;Moreno-Chamarro, Ortega, et al., 2017;Moreno-Chamarro, Zanchettin, Lohmann, Luterbacher, et al., 2017;Thirumalai et al., 2018). Whatever the underlying causes for the LIA, the Harvey Lake core record presents evidence that there was considerable variability in runoff during this turbulent climatic interval and the Br + Cl/Al data indicates a strong marine air mass influence across the region. ...
Major Tropical Cyclone (TC) events cause extensive damage in coastal regions of the western North Atlantic Basin. The short instrumental record leaves significant gaps in understanding long‐term trends in TC recurrence and intensity, creating uncertainty about future storm trends. Analysis of an ∼520‐year core record from Harvey Lake, located >80 km from the Atlantic coast in southwestern New Brunswick, Canada was carried out using: (a) end‐member mixing analysis (EMMA) of lake sediment grain size data to identify storm‐linked sedimentological processes; and (2) ITRAX X‐ray fluorescence (XRF) derived element/ratios (Fe, Ti, Ca/Sr, Zr/Rb, K/Rb, and Br + Cl/Al) associated with precipitation, weathering, catchment runoff, and air masses. Three derived end members were correlated to heavy rainfall events (EM01), spring freshet (EM02), and TCs (EM03). CONISS analysis of the EMMA and XRF core data resulted in recognition of four unique climatic zones distinguished by distinct distributions of TC and rainfall/weathering/runoff/and air masses. Numerous, major (EM01) rainfall events and (EM03) TC events characterized the basal core record during the early Little Ice Age (LIAa; Zone 1) phase, terminating at ∼1645. A near cessation of heavy rainfall and TC events differentiated the subsequent colder LIAb (∼1645–1825; Zone 2) and subsequent Little Ice Age Transition (∼1825–1895; Zone 3). A resurgence of major rainfall and TC events occurred during recovery from the LIA starting in ∼1895 (Zone 4). EMMA provides a robust tool for recognition of TC and major rainfall events, and greatly expands the potential for paleo‐storm activity research well inland from coastal regions.
... Thirumalai et al., 2018) is depicted with a yellow star and the location of the sediment trap (Richey et al., 2019) is indicated with an inverted yellow triangle. Observed SST and SSS data were taken from HadISST (Rayner et al., 2003) and ORA-S4 (Balmaseda et al., 2012) data sets, respectively. ...
... Well-resolved reconstructions of Holocene climate change can establish important baselines to separate forced responses from internal variability and can inform comparisons with climate simulations. Although there are some high-resolution geochemical records concentrated in the late Holocene (Lund & Curry, 2006;Richey et al., 2007Richey et al., , 2009Thirumalai et al., 2018) and a few in the early Holocene (LoDico et al., 2006;Schmidt et al., 2012), continuous and quantitative reconstructions of surface-ocean conditions in the GoM over the full epoch are lacking. This deficit has restricted efforts to delineate mechanisms involved in regional paleoclimate changes as well as shifts in low-latitude circulation across the Atlantic (Metcalfe et al., 2015). ...
... We then use our new data set. alongside replicated late Holocene reconstructions from the same basin (see Thirumalai et al., 2018) to create a composite record spanning the past 11,700 years. We then synthesize the available records across the GoM to investigate spatiotemporal trends over the Holocene. ...
Flows into and out of the Gulf of Mexico (GoM) are integral to North Atlantic Ocean circulation and help facilitate poleward heat transport in the Western Hemisphere. The GoM also serves as a key source of moisture for most of North America. Modern patterns of sea‐surface temperature (SST) and salinity in the GoM are influenced by the Loop Current, its eddy‐shedding dynamics, and the ensuing interplay with coastal processes. Here, we present sub‐centennial‐scale records of SST and stable oxygen isotope composition of seawater (δ¹⁸Osw; a proxy for salinity) over the past 11,700 years using planktic foraminiferal geochemistry in sediments from the Garrison Basin, northwestern GoM. We measured δ18O and magnesium‐to‐calcium ratios in tests of Globigerinoides ruber (white) to generate quantitative estimates of past sea‐surface conditions. Our results replicate and extend late Holocene reconstructions from the Garrison Basin, using which we then create composites of SST and δ¹⁸Osw. We find considerable centennial and millennial‐scale variability in both SST and δ¹⁸Osw, although their evolution over the Holocene is distinct. Whereas mean‐annual SSTs display pronounced millennial‐scale variability, δ¹⁸Osw exhibits a secular trend spanning multiple millennia and points to increasing northwestern GoM surface salinity since the early Holocene. We then synthesize the available Holocene records from across the GoM and alongside the Garrison Basin composite uncover substantial, yet regionally consistent, spatiotemporal variability. Finally, we discuss the role of the Loop Current and coastal influx of freshwater in imposing these heterogeneities. We conclude that dynamic surface‐ocean changes occurred across the GoM over the Holocene.
... The δ 18 O of foraminiferal shells is dependent on two main factors: (i) the temperature at which the oxygen isotopic fractionation between the foraminiferal shell and seawater occurred (Urey, 1948;Emiliani, 1954aEmiliani, , 1954b and (ii) the δ 18 O of seawater at the time of calcification (Olausson, 1965;Shackleton, 1967;Labeyrie et al., 1992). This dependence of foraminiferal shell δ 18 O on the isotopic composition of palaeo-seawater is utilized to understand past ocean dynamics as the isotopic composition of ocean surface water (δ 18 O sw ) changes in response to various physical processes such as evaporation (E), precipitation (P), runoff (R) (Craig and Gordon, 1965;Clark and Fritz, 1999), advection (Friedman et al., 2017;Thirumalai et al., 2018) and the glacial ice stored in continental ice sheets (ice volume effect) (Olausson, 1965;Shackleton, 1967). δ 18 O sw increases with increasing rate of evaporation, and on the other hand, decreases with precipitation as well as the influx of continental runoff. ...
Palaeosalinity (S*) estimates rely on calculations stemming from the oxygen isotopic composition of foraminiferal shells (δ¹⁸OF) preserved in ocean sediments. δ¹⁸OF depends on palaeo in-situ temperature (T*) and the oxygen isotopic composition of seawater (δ¹⁸Osw), which in turn depends on various physical oceanographic conditions including evaporation, precipitation, advection, and freshwater fluxes. δ¹⁸Osw and salinity covary linearly owing to similar controls, as they both increase with increases in evaporation and decrease with higher precipitation and continental runoff. Typically, palaeosalinity is estimated assuming a constant spatial and temporal relationship between δ¹⁸Osw and salinity. However, evaporative and freshwater fluxes exhibit spatiotemporal variability and thus, can change the slope and intercept of the linear δ¹⁸O-salinity relationship. Hence, the use of a constant δ¹⁸Osw-salinity relationship may produce (systematic) errors in palaeosalinity estimates. Here, we revisit palaeosalinity estimates and provide a reassessment of errors arising from these calculations based on recently reported δ¹⁸Osw-salinity relationships in the northern Indian Ocean. Our calculations point to errors in palaeosalinity that can be as large as 55% - significant enough to change our understanding of past ocean circulation and palaeomonsoon variability. These results have important implications for reconstructing the palaeoceanographic record and interpreting past ocean-atmosphere variability.
... However, terrestrial paleoclimate proxies like tree rings do not directly record the atmospheric pressure or wind anomalies associated with the NAO nor the SSTs used to characterize the AMO but can reflect these indirectly via their influence on precipitation or temperature at the location of the tree-ring chronology. Marine records from sediments or corals may be used to reconstruct SSTs, but only some have the temporal resolution, age model precision, and length to do so at decadal time scales over the last millennium in the North Atlantic itself (Moffa-Sánchez et al., 2014a;Moffa-Sánchez et al., 2014b;Moffa-Sánchez et al., 2019;Thirumalai et al., 2018). A common approach to reconstructing the NAO or AMO is therefore to directly target indices of these modes as predictands in a statistical regression model where tree-ring chronologies are the potential predictors (cf. ...
... These years are associated with a wetter Mediterranean and drier Northern Europe and show stronger and southerly displaced upper level (250 mb) westerly winds. Conversely, dry years in the Mediterranean and Thirumalai et al., 2018). In the Atlantic, the pattern is similar to the regression of the SST field on the AMO index (Figure 4b), including the large anomalies in the Labrador Sea and weaker anomalies in the western subtropical Atlantic. ...
... Ocean sediment cores suggest the early LIA was characterized by a weakened Gulf Stream, reduced Atlantic surface-ocean circulation, and colder SSTs (Lund et al., 2006;Moffa-Sánchez et al., 2014a, 2014b, Thirumalai et al., 2018. While model simulations show that an abrupt weakening of the subpolar gyre can cause LIA conditions in the Atlantic (Moreno-Chamarro et al., 2016), in agreement with ocean sediment proxies (Moffa-Sánchez et al., 2014b), models are inconsistent on the forced response of the gyre (e.g., Lehner et al., 2013;Schleussner & Feulner, 2013). ...
Hydroclimate extremes in North America, Europe, and the Mediterranean are linked to ocean and atmospheric circulation anomalies in the Atlantic, but the limited length of the instrumental record prevents complete identification and characterization of these patterns of covariability especially at decadal to centennial timescales. Here we analyze the coupled patterns of drought variability on either side of the North Atlantic Ocean basin using independent climate field reconstructions spanning the last millennium in order to detect and attribute epochs of coherent basin-wide moisture anomalies to ocean and atmosphere processes. A leading mode of broad-scale moisture variability is characterized by distinct patterns of North Atlantic atmosphere circulation and sea surface temperatures. We infer a negative phase of the North Atlantic Oscillation and colder Atlantic sea surface temperatures in the middle of the 15th century, coincident with weaker solar irradiance and prior to strong volcanic forcing associated with the early Little Ice Age.
