As photovoltaic power is expanding rapidly worldwide, it is imperative to assess its promise under future climate scenarios. While a great deal of research has been devoted to trends in mean solar radiation, less attention has been paid to its intermittent character, a key challenge when compounded with uncertainties related to climate variability. Using both satellite data and climate model outputs, we characterize solar radiation intermittency to assess future photovoltaic reliability. We find that the relation between the future power supply and long-term mean solar radiation trends is spatially heterogeneous, showing power reliability is more sensitive to the fluctuations of mean solar radiation in hot arid regions. Our results highlight how reliability analysis must account simultaneously for the mean and intermittency of solar inputs when assessing the impacts of climate change on photovoltaics.

Clouds’ efficiency at reflecting solar radiation and trapping the terrestrial radiation is strongly modulated by the diurnal cycle of clouds (DCC). Much attention has been paid to mean cloud properties due to their critical role in climate projections; however, less research has been devoted to the DCC. Here we quantify the mean, amplitude, and phase of the DCC in climate models and compare them with satellite observations and reanalysis data. While the mean appears to be reliable, the amplitude and phase of the DCC show marked inconsistencies, inducing overestimation of radiation in most climate models. In some models, DCC appears slightly shifted over the ocean, likely as a result of tuning and fortuitously compensating the large DCC errors over the land. While this model tuning does not seem to invalidate climate projections because of the limited DCC response to global warming, it may potentially increase the uncertainty of climate predictions.

The local role that land‐atmosphere interactions play in the rainfall process has been often explored by investigating the initiation of moist convection as the top of the atmospheric boundary layer (ABL) crosses the lifting condensation level (LCL). However, this LCL crossing alone is not a sufficient indicator of the probability and intensity of subsequent convective precipitation, which is instead better characterized by the added consideration of the so‐called convective available potential energy (CAPE). In this study, both the LCL crossing and CAPE are jointly considered as the primary indicators of the occurrence and intensity of moist convection in order to analyze the land‐atmosphere interactions through a simple soil‐plant system and a zero‐dimensional mixed‐layer model. The approach is explored using the free atmospheric conditions observed at the Central Facility in the Southern Great Plains, where the ABL analysis shows both dry and wet soil can be conducive to early moist convection depending on atmospheric conditions but CAPE always tends to be larger under wetter soil conditions. The combination of the two indicators, LCL crossing and CAPE, further allows us to classify free atmosphere and soil moisture regimes into positive and negative feedback regimes for moist convection.

Land‐surface partitioning of net radiation into sensible and latent heat fluxes is critical for hydroclimatic processes but remains highly uncertain due to limited observations. We show that a suitable extension of the Budyko's curve, a well‐known framework in hydrology for water balance estimation, can be utilized effectively to partition the surface energy fluxes by expressing the long‐term evaporative fraction (EF) as a function of the dryness index only. The combination of this energy partitioning method with hydrological observations allows us to estimate the surface energy components at watershed and continental scales. Using this new framework, we show that North American Regional Reanalysis data overestimate surface evaporation, likely influencing the modeling of atmospheric convection. The obtained hydrologic constrains on energy partitioning can be used to provide more accurate estimations of surface energy fluxes for hydroclimatic predictions.

The soil moisture‐rainfall feedback (SMRF) may significantly impact hydro‐climatic dynamics, inducing persistent weather conditions that are responsible for prolonged droughts or abnormally wet states. However, externally driven seasonal variability in rainfall and potential evapotranspiration, with the associated patterns of wet and dry conditions, may both interact with such an SMRF. In this study, seasonal variations in radiation and precipitation forcing are included in a stochastic SMRF model with the assumption of a soil moisture‐dependent average rainfall frequency to explore their effects on the soil moisture probabilistic structure. The theoretical model results, based on a parameterization using data for soil moisture and climate in Illinois, show that average rainfall frequency peaks in late spring when both the soil condition and the SMRF strength favor convective rainfall triggering. Under such conditions, the soil moisture tends to exhibit bimodal behavior until the SMRF strength becomes weak again toward the end of the growing season. Such a behavior is reminiscent of the dynamics of a system undergoing a periodic, stochastically forced pitchfork bifurcation. The presence of bimodal soil moisture behavior is also verified using nonparametric statistical tests on soil moisture data. The analysis of wet‐to‐wet and dry‐to‐dry soil moisture transitions in the joint probability distribution of soil moisture further corroborates the presence of hydro‐climatic persistence in the spring‐to‐summer transition.