Dynamic and thermodynamic influences on precipitation in Northeast Mexico on orbital to millennial timescales

Abstract

The timing and mechanisms of past hydroclimate change in northeast Mexico are poorly constrained, limiting our ability to evaluate climate model performance. To address this, we present a multiproxy speleothem record of past hydroclimate variability spanning 62.5 to 5.1 ka from Tamaulipas, Mexico. Here we show a strong influence of Atlantic and Pacific sea surface temperatures on orbital and millennial scale precipitation changes in the region. Multiple proxies show no clear response to insolation forcing, but strong evidence for dry conditions during Heinrich Stadials. While these trends are consistent with other records from across Mesoamerica and the Caribbean, the relative importance of thermodynamic and dynamic controls in driving this response is debated. An isotope-enabled climate model shows that cool Atlantic SSTs and stronger easterlies drive a strong inter-basin sea surface temperature gradient and a southward shift in moisture convergence, causing drying in this region.

Fig. 1. Summer (JJAS) climatology and nearby paleoclimate records.

Map of regional precipitation and magnitude of low-level (850 mb) winds using PERSIANN precipitation data and winds from MPI-ESM-Historical. Nearby records include (1) an ocean sediment core from the Gulf of Mexico and the (2) Florida Straight, speleothem records from (3) Cuba, (4) Southern Mexico and (8) Costa Rica, additional ocean sediment cores from the (6) Caribbean Sea, and (7) Cariaco Basin, and a lake sediment core (5) from Guatemala.

Fig. 2. Stalagmite CB2, age-depth model and Mg/Ca, δ¹⁸O and δ¹³C results.

a Results of 1578 stable isotope and 789 trace element measurements. δ¹⁸O is top and blue, δ¹³C is central and in orange, Mg/Ca is in bottom and green. Dates with associated uncertainties are below Mg/Ca. Heinrich Stadials highlighted in light red. b CB2 Age-Depth Model constructed using 2000 Monte-Carlo simulations via the age-depth modeling software COPRA and 33 U-Th ages. Uncertainty in age-depth model indicated by gray shading, uncertainty in U-Th ages indicated by red error bars. c Sample CB2 after being cut and polished.

Fig. 3. Comparison of CB2 δ¹⁸O to various potential forcings.

a Autumn (SON, r = −0.48, p < 0.14, orange) and Summer (JJA, r = −0.26 p = 0.46 black) insolation. b Atmospheric pCO₂ (r = −0.61, p < 0.01, maroon). c SSTs from Gulf of Mexico (r = −0.52, p < 0.03, silver), Caribbean (r = −0.59, p < 0.05, lime green), Tropical N. Pacific (r = −0.55, p < 0.03, magenta) and Tropical N. Atlantic (r = −0.73, p < 0.01, teal). d Greenland temperatures (r = −0.61, p < 0.01, indigo).

Fig. 4. iCESM1 simulation of Heinrich Stadials compared to LGM over North America.

a Annual temperature and Sea Level Pressure changes. b Annual precipitation change, with statistically significant changes contoured by dashed gray lines. c Annual changes in total column precipitable water and 850 mb winds. d Annual changes in the stable oxygen isotope ratio of precipitation.

Fig. 5. Comparison of CB2 with other paleoclimate records.

Comparison of CB2 δ¹³O (blue) and δ¹³C (orange) to Pa/Th ratios (burgundy; Henry et al. ) with Florida Strait gulf stream circulation (magenta), Cariaco Basin reflectance (purple), Juxtlahuaca (light green), Costa Rica (blue), Cuba (dark green) and Lake Petén-Itzá (black).