| 22 Oct 2025
Improving Arctic Surface Radiation Estimation Using a Nonlinear Perturbation Model with a Fused Multi-Satellite Cloud Fraction Dataset
Abstract. Arctic has been undergoing rapid climate change, where radiative processes are key controlling factors. However, cloud-related uncertainties remain the primary barrier to accurately estimating the radiation budget. Strong coupling between clouds and other variables complicates the isolation of cloud-related impacts, and the linear assumptions in traditional models further restrict attribution of radiation changes to cloud-related influences. This study introduces an artificial neural network model that emulates radiative components typically represented in radiative transfer or climate models. Without relying on linear assumptions, the model directly quantifies the influence of cloud fraction (CF) on radiation. Using a more accurate CF dataset, we refined the monthly downwelling shortwave radiation (DSR) estimates from Clouds and the Earth's Radiant Energy System (CERES) SYN products and further estimated all-wave net radiation (NR) from the corrected DSR. Validation against ground-based observations confirmed that the CF-corrected DSR effectively mitigated the overestimation in CERES DSR, reducing biases by up to 23 W m⁻². At sites where CF underestimation exceeded 25 %, the monthly-mean bias decreased from 25.70 W/m² to 4.88 W/m², with RMSE reduced from 40.36 W/m² to 32.60 W/m². The estimated monthly NR also improved markedly (RMSE reduced from 34.88 W/m² to 28.90 W/m²). Under large CF underestimation (>30 %), the CERES NR nearly failed (R² = 0.0182), whereas NR derived from CF-corrected DSR retained reasonable agreement (R² = 0.5411). Importantly, this work produces a new NR dataset with enhanced accuracy over the Arctic, offering direct value for studies of surface energy balance, climate feedbacks, and long-term variability.
This manuscript introduces a new neural network (NN) based approach for correcting estimates of Arctic downwelling shortwave radiation (DSR) and, subsequently, Arctic net surface radiation (NR) from CERES products, cloud properties, and ancillary atmospheric and surface properties. The algorithm improves DSR estimates relative to surface flux observations, especially in cases where CERES underestimates cloud fraction relative to a recently developed cloud product that combines active and passive observations. NR is also improved, though to a lesser degree, likely due to the approach neglecting the varying influences of downwelling longwave radiation (DLR) on NR. There is value to producing more robust DSR estimates that incorporate improved estimates of cloud fraction from active sensors as well as dependences on other atmospheric and surface conditions.
My primary concern with the study is that the NN approach introduces a disconnect between the final DSR and NR estimates and the physics that modulated them. While the multi-variate NN captures nonlinear relationships and includes additional factors that modulate surface radiative fluxes, it masks the precise physical relationships that led to the results. One clear example of this is the fact that NR is estimated from DSR without accounting for cloud or atmospheric influences on longwave radiation. Presumably the NN captures some of the longwave effects through covariances between DSR, DLR, and other regression variables but it cannot make up for the lack of information provided by longwave radiative transfer calculations. The direct physical connection between inputs and simulated fluxes has considerable value for many atmospheric process and climate applications, so it is not clear how this product could be used in those contexts.
Considering both the value of the analysis and the associated concerns, the paper may be suitable for publication after major revisions to better explain which applications the NR product may be address and responding to the following comments.
Specific Comments: