Natural Marine Precursors Boost Continental New Particle Formation and Production of Cloud Condensation Nuclei.

Marine dimethyl sulfide (DMS) emissions are the dominant source of natural sulfur in the atmosphere. DMS oxidizes to produce low-volatility acids that potentially nucleate to form particles that may grow into climatically important cloud condensation nuclei (CCN). In this work, we utilize the chemistry transport model ADCHEM to demonstrate that DMS emissions are likely to contribute to the majority of CCN during the biological active period (May-August) at three different forest stations in the Nordic countries. DMS increases CCN concentrations by forming nucleation and Aitken mode particles over the ocean and land, which eventually grow into the accumulation mode by condensation of low-volatility organic compounds from continental vegetation. Our findings provide a new understanding of the exchange of marine precursors between the ocean and land, highlighting their influence as one of the dominant sources of CCN particles over the boreal forest.

Fire frequency drives decadal changes in soil carbon and nitrogen and ecosystem productivity.

Fire frequency is changing globally and is projected to affect the global carbon cycle and climate. However, uncertainty about how ecosystems respond to decadal changes in fire frequency makes it difficult to predict the effects of altered fire regimes on the carbon cycle; for instance, we do not fully understand the long-term effects of fire on soil carbon and nutrient storage, or whether fire-driven nutrient losses limit plant productivity. Here we analyse data from 48 sites in savanna grasslands, broadleaf forests and needleleaf forests spanning up to 65 years, during which time the frequency of fires was altered at each site. We find that frequently burned plots experienced a decline in surface soil carbon and nitrogen that was non-saturating through time, having 36 per cent (±13 per cent) less carbon and 38 per cent (±16 per cent) less nitrogen after 64 years than plots that were protected from fire. Fire-driven carbon and nitrogen losses were substantial in savanna grasslands and broadleaf forests, but not in temperate and boreal needleleaf forests. We also observe comparable soil carbon and nitrogen losses in an independent field dataset and in dynamic model simulations of global vegetation. The model study predicts that the long-term losses of soil nitrogen that result from more frequent burning may in turn decrease the carbon that is sequestered by net primary productivity by about 20 per cent of the total carbon that is emitted from burning biomass over the same period. Furthermore, we estimate that the effects of changes in fire frequency on ecosystem carbon storage may be 30 per cent too low if they do not include multidecadal changes in soil carbon, especially in drier savanna grasslands. Future changes in fire frequency may shift ecosystem carbon storage by changing soil carbon pools and nitrogen limitations on plant growth, altering the carbon sink capacity of frequently burning savanna grasslands and broadleaf forests.

Automating Wood Species Detection and Classification in Microscopic Images of Fibrous Materials with Deep Learning.

We have developed a methodology for the systematic generation of a large image dataset of macerated wood references, which we used to generate image data for nine hardwood genera. This is the basis for a substantial approach to automate, for the first time, the identification of hardwood species in microscopic images of fibrous materials by deep learning. Our methodology includes a flexible pipeline for easy annotation of vessel elements. We compare the performance of different neural network architectures and hyperparameters. Our proposed method performs similarly well to human experts. In the future, this will improve controls on global wood fiber product flows to protect forests.

Process-evaluation of forest aerosol-cloud-climate feedback shows clear evidence from observations and large uncertainty in models.

Natural aerosol feedbacks are expected to become more important in the future, as anthropogenic aerosol emissions decrease due to air quality policy. One such feedback is initiated by the increase in biogenic volatile organic compound (BVOC) emissions with higher temperatures, leading to higher secondary organic aerosol (SOA) production and a cooling of the surface via impacts on cloud radiative properties. Motivated by the considerable spread in feedback strength in Earth System Models (ESMs), we here use two long-term observational datasets from boreal and tropical forests, together with satellite data, for a process-based evaluation of the BVOC-aerosol-cloud feedback in four ESMs. The model evaluation shows that the weakest modelled feedback estimates can likely be excluded, but highlights compensating errors making it difficult to draw conclusions of the strongest estimates. Overall, the method of evaluating along process chains shows promise in pin-pointing sources of uncertainty and constraining modelled aerosol feedbacks.