This study aims to investigate the global hydrologic cycle during the mid-Cretaceous greenhouse by utilizing the oxygen isotopic composition of pedogenic carbonates (calcite and siderite) as proxies for the oxygen isotopic composition of precipitation. The data set builds on the Aptian–Albian sphaerosiderite δ¹⁸O data set presented by Ufnar et al. (2002) by incorporating additional low latitude data including pedogenic and early meteoric diagenetic calcite δ¹⁸O. Ufnar et al. (2002) used the proxy data derived from the North American Cretaceous Western Interior Basin (KWIB) in a mass balance model to estimate precipitation–evaporation fluxes. We have revised this mass balance model to handle sphaerosiderite and calcite proxies, and to account for longitudinal travel by tropical air masses. We use empirical and general circulation model (GCM) temperature gradients for the mid-Cretaceous, and the empirically derived δ¹⁸O composition of groundwater as constraints in our mass balance model. Precipitation flux, evaporation flux, relative humidity, seawater composition, and continental feedback are adjusted to generate model calculated groundwater δ¹⁸O compositions (proxy for precipitation δ¹⁸O) that match the empirically-derived groundwater δ¹⁸O compositions to within ± 0.5‰. The model is calibrated against modern precipitation data sets.

Four different Cretaceous temperature estimates were used: the leaf physiognomy estimates of Wolfe and Upchurch (1987) and Spicer and Corfield (1992), the coolest and warmest Cretaceous estimates compiled by Barron (1983) and model outputs from the GENESIS-MOM GCM by Zhou et al. (2008). Precipitation and evaporation fluxes for all the Cretaceous temperature gradients utilized in the model are greater than modern precipitation and evaporation fluxes. Balancing the model also requires relative humidity in the subtropical dry belt to be significantly reduced. As expected calculated precipitation rates are all greater than modern precipitation rates. Calculated global average precipitation rates range from 371 mm/year to 1196 mm/year greater than modern precipitation rates. Model results support the hypothesis that increased rainout produces δ¹⁸O-depleted precipitation.

Sensitivity testing of the model indicates that the amount of water vapor in the air mass, and its origin and pathway, significantly affect the oxygen isotopic composition of precipitation. Precipitation δ¹⁸O is also sensitive to seawater δ¹⁸O and enriched tropical seawater was necessary to simulate proxy data (consistent with fossil and geologic evidence for a warmer and evaporatively enriched Tethys). Improved constraints in variables such as seawater δ¹⁸O can help improve boundary conditions for mid-Cretaceous climate simulations.