A multivariate probability density function (PDF) can be used to represent the subgrid (below grid-box size) variability of atmospheric fields. The PDF was previously extended to include hydrometeor fields, such as rain water mixing ratio. Now, the PDF of hydrometeor fields is altered to account for precipitating and precipitation-less regions of the subgrid domain. Accounting for these regions allowed the hydrometeor PDF to produce an improved match to results from large-eddy simulations.

Taking into account model inadequacies is a key challenge in climate forecasting. As part of the FP7-SPECS project, we examine how stochastic perturbations of atmospheric model dynamics impact seasonal forecast quality of the CNRM coupled model. The method described in this paper helps derive model error statistics as well as improve key aspects of our forecasting system such as systematic errors over the North Atlantic mid-latitudes.

In order to evaluate climate models, scientists have historically run them using prescribed (for example observed) sea surface temperatures; however, no such restriction is typically applied to the land. This study presents a method of prescribing the land temperatures in a climate model and shows that the resultant climate simulation is consistent with the free running simulation. Such a model will be useful for perturbing and fixing surface temperatures globally, as demonstrated in this paper.

We propose a butterfly implementation for data transfer. Although the butterfly implementation outperforms the existing implementation (the P2P implementation) in many cases, it degrades the performance in some cases. So we design and implement an adaptive data transfer library that automatically chooses an optimal implementation between the P2P one and the butterfly one and also further improves the performance based on the butterfly implementation through skipping some butterfly stages.

The present work aims at evaluating the scalability performance of a high-resolution global ocean biogeochemistry model (PELAGOS025) on massive parallel architectures and the benefits in terms of the time-to-solution reduction. The outcome of the analysis demonstrated that the lack of scalability is due to several factors such as the I/O operations, the memory contention, and the load unbalancing due to the memory structure of the biogeochemistry model component.

In its purest essence, numerical weather prediction boils down to solving the fundamental laws of nature with computers. Such fundamental laws are the conservation of energy and the conservation of mass. In this paper, a framework is presented that allows to respect these laws more accurately, which should lead to weather forecasts that correspond better to reality. Under specific circumstances, such as heavy precipitation, the proposed framework has a significant impact on the forecast.

The atmospheric oxidation of aromatic compounds is an important source of aerosol-forming species, and thus contributes to pollution in urban areas. However, details of the mechanisms by which oxidation occurs are only recently being elucidated. Here we report the incorporation of a newly developed mechanism for aromatic oxidation into the UCI-CIT regional air quality model. Results suggest an unexpected role for chemical pathways typically associated with cleaner environments.

The aerosol adjoint module of the atmospheric chemical modeling system GRAPES–CUACE was developed, tested for its correctness, and used in a receptor–source sensitivity test. The results showed that controlling critical emission sources during critical time intervals on the basis of adjoint sensitivity analysis is much more efficient than controlling administrative specified regions during an experiential time period.

With the availability of lake surface water temperature (LSWT) satellite data for 246 globally distributed large lakes, we tune a lake model, FLake, by varying 3 basic lake properties, shown to have the most influence over the modelled LSWTs. Tuning reduces the mean absolute difference (between model and satellite LSWTs) from an average of 3.38 ºC per day (untuned model) to 0.85 ºC per day (tuned model). The effect of several LSWT drivers, such as wind speed and lake depth are also demonstrated.

Biogenic volatile organic compounds (BVOCs) participate in reactions that can lead to secondarily formed ozone and particulate matter impacting air quality and climate and are important inputs for atmospheric models. BVOC emissions are sensitive to the vegetation species and leaf temperature. Here, we have improved the vegetation data and modeled leaf temperature of the Biogenic Emission Inventory System model. Updated algorithms improved model evaluation against observations in California.

Four new radiation related submodels (RAD, AEROPT, CLOUDOPT, and ORBIT) are available within the MESSy framework now. They are largely based on the original radiation scheme of ECHAM5. RAD simulates radiative transfer, AEROPT calculates aerosol optical properties, CLOUDOPT calculates cloud optical properties, and ORBIT is responsible for Earth orbit calculations. Multiple diagnostic calls of the radiation routine are possible, so radiative forcing can be calculated during the model simulation.

mizuRoute version 1 is a stand-alone runoff routing tool that post-processes runoff outputs from any distributed hydrologic models to produce streamflow estimates in large-scale river network. mizuRoute is flexible to river network representation and includes two different river routing schemes. This paper demonstrates mizuRoute's capability of multi-decadal streamflow estimations in the river networks over the entire contiguous Unites States, which contains over 54 000 river segments.

A complex thermodynamic sea ice model (HIGHTSI) is coupled with WRF to improve its simulation of surface temeprature and surface upward long-wave radiation flux over sea ice surface. HIGHTSI includes the ablation and accretion processes of sea ice and uses an interpolation method which can ensure the heat conservation during its integration. These allow the HIGHTSI to better resolve the energy balance in the sea ice, and the bias in sea ice temperature is reduced considerably.

Recent decades have seen significant developments in seasonal-to-interannual timescale climate prediction. However, until recently the potential of such systems to predict Arctic climate had not been assessed. This paper describes a multi-model predictability experiment which was run as part of the Arctic Predictability and Prediction On Seasonal to Interannual Timescales (APPOSITE) project. The main goal of APPOSITE was to quantify the timescales on which Arctic climate is predictable.

We present a redesign and further development of a complex and well-known aquatic ecosystem model (PCLake) into the Framework for Aquatic Biogeochemical Models (FABM). So PCLake can run in different hydrodynamic environments, ranging from 0-D to 3-D. We introduce the methods and technical details about how the model was re-designed into a modular structure and the new features of PCLake enabled by FABM. We further present a benchmark test case to verify the new model implementation.

This paper presents the first known application of multi-model ensembles to the forecasting of the thermosphere. A multi-model ensemble (MME) is a method for combining different, independent models. The main advantage of using an MME is to reduce the effect of model errors and bias, since it is expected that the model errors will, at least partly, cancel. This paper shows that use of MMEs for forecasting thermospheric densities can reduce errors by 60 %.