Next Article in Journal
Potential Climate Impacts of Hydrological Alterations and Discharge Variabilities of the Mura, Drava, and Danube Rivers on the Natural Resources of the MDD UNESCO Biosphere Reserve
Previous Article in Journal
Intensity, Duration and Spatial Coverage of Aridity during Meteorological Drought Years over Northeast Thailand
 
 
Search for Articles:
Title / Keyword
Author / Affiliation / Email
Journal
Article Type
 
 
Advanced Search
 
Section
Special Issue
Volume
Issue
Number
Page
 
Journals
Climate
Volume 10
Issue 10
10.3390/cli10100138
climate-logo
Submit to this Journal Review for this Journal Propose a Special Issue
► Article Menu

Article Menu

share Share announcement Help format_quote Cite question_answer Discuss in SciProfiles thumb_up
...
Endorse textsms
...
Comment
first_page
settings
Order Article Reprints
Font Type:
Arial Georgia Verdana
Font Size:
Aa Aa Aa
Line Spacing:
Column Width:
Background:
Article

On the Intercontinental Transferability of Regional Climate Model Response to Severe Forestation

by
Olivier Asselin
1,*,
Martin Leduc
1,
Dominique Paquin
1,
Alejandro Di Luca
2,
Katja Winger
2,
Melissa Bukovsky
3,
Biljana Music
1 and
Michel Giguère
1
1
Ouranos, Montréal, QC H3A 1B9, Canada
2
Centre ESCER, Université du Québec à Montréal, Montréal, QC H3C 3P8, Canada
3
National Center for Atmospheric Research, Boulder, CO 80305, USA
*
Author to whom correspondence should be addressed.
Climate 2022, 10(10), 138; https://doi.org/10.3390/cli10100138
Submission received: 11 August 2022 / Revised: 7 September 2022 / Accepted: 13 September 2022 / Published: 23 September 2022
(This article belongs to the Section Climate Dynamics and Modelling)
Browse Figures
Review Reports Versions Notes

Abstract

:
The biogeophysical effects of severe forestation are quantified using a new ensemble of regional climate simulations over North America and Europe. Following the protocol outlined for the Land-Use and Climate Across Scales (LUCAS) intercomparison project, two sets of simulations are compared, FOREST and GRASS, which respectively represent worlds where all vegetation is replaced by trees and grasses. Three regional climate models were run over North America. One of them, the Canadian Regional Climate Model (CRCM5), was also run over Europe in an attempt to bridge results with the original LUCAS ensemble, which was confined to Europe. Overall, the CRCM5 response to forestation reveals strong inter-continental similarities, including a pronounced wintertime and springtime warming concentrated over snow-masking evergreen forests. Crucially, these northern evergreen needleleaf forests populate lower, hence sunnier, latitudes in North America than in Europe. Snow masking reduces albedo similarly over both continents, but stronger insolation amplifies the net shortwave radiation and hence warming simulated over North America. In the summertime, CRCM5 produces a mixed response to forestation, with warming over northern needleleaf forests and cooling over southern broadleaf forests. The partitioning of the turbulent heat fluxes plays a major role in determining this response, but it is not robust across models over North America. Implications for the inter-continental transferability of the original LUCAS results are discussed.

1. Introduction

Afforestation and reforestation, herein combined as forestation, could remove significant amounts of carbon dioxide from the atmosphere [1,2]. Large-scale forestation, however, would also alter energy and water exchanges between the land and the atmosphere [3]. Turning grassland into forests, for instance, may lower albedo, increase surface roughness, and facilitate the pumping of water from the soil to the atmosphere. These effects are collectively known as biogeophysical effects, in contrast to the biogeochemical effects pertaining to greenhouse gases and aerosol precursors [4]. While forestation leads to biogeochemical cooling via carbon sequestration, biogeophysical effects may cause warming or cooling depending on a variety of factors such as latitude, time of year, tree species, and original land cover [5,6]. With forestation expected to contribute about a quarter of mitigation efforts pledged under the Paris agreement, it is essential that biophysical effects are quantified and accounted for.
The Land-Use and Climate, Identification of Robust Impacts (LUCID) project was the first global climate model intercomparison effort aiming to quantify the biogeophysical effects of historical deforestation [7,8,9]. While biogeophysical effects of historical deforestation were found to be modest when averaged over the globe, they could nevertheless match the magnitude of the more spatially diffuse biogeochemical effects at the regional scale. However, the biogeophysical effects of such land-use changes are not routinely included in regional climate model (RCM) intercomparisons [10,11,12]. Numerous single-RCM studies have investigated the biogeophysical effects of land-use changes [13,14,15,16,17], but the lack of a shared protocol makes comparison between models difficult.
The Land-Use and Climate Across Scales (LUCAS) initiative is a flagship pilot study (FPS) of the World Climate Research Program-Coordinated Regional Climate Downscaling Experiment (WCRP-CORDEX) designed to improve the integration of land-use change in RCMs and to quantify their biogeophysical effects on climate [18]. The first phase of LUCAS focuses on the biogeophysical effects of severe forestation in Europe using an ensemble of nine combinations of RCMs and land surface models [19]. For every model combination, two sets of simulations are compared, FOREST and GRASS, respectively representing worlds where all vegetation is replaced with trees and grasses. This large simulation ensemble allowed investigations into various effects of forestation such as snow cover [20,21], the diurnal air temperature cycle [22], the seasonal soil temperature cycle [23], surface roughness and its role in the partition of turbulent heat fluxes [24], and the land–atmosphere coupling [25,26].
Are LUCAS findings specific to Europe, or may they be applicable to North America as well? To address this question, we apply the protocol designed for the original European LUCAS initiative to a new set of RCMs and a new continent: North America. One of the ensemble members—version 5 of the Canadian Regional Climate Model (CRCM5; [27,28])—is also run over Europe. In the following section, the experiment setup is described in more detail. Then, CRCM5 simulations are compared over North America and Europe, providing a unique look at the transferability of LUCAS findings to North America (Section 3.1). In Section 3.2, the robustness of the intercontinental transferability is assessed by comparing all three models of the present ensemble over North America. The various components of land–atmosphere energy fluxes are then averaged over the main forest families for both North America and Europe (Section 3.3). Implications are discussed in Section 4.

2. Materials and Methods

2.1. Model Ensemble

This paper presents simulations from three combinations of RCMs and land surface models, the main properties of which are outlined in Table 1. The Weather and Research Forecast (WRF; [29]) model version 3.5.1 is coupled to the Unified NOAH land surface model. Two versions of the Canadian Regional Climate Model (CRCM5 and CRCM6) are also employed, each coupled to a slightly different version of the Canadian Land Surface Scheme (CLASS; [30,31]). Differences between the two versions of CLASS—revised ponding depth over organic soils, revised snow albedo refreshment threshold, and new snow thermal conductivity algorithm—are modest in comparison to those between the CRCM5 and CRCM6 atmospheric components. The CRCMs are built on two different versions of the Global Environmental Multiscale Model dynamical core: 3.3.3.1 for CRCM5 [32] and 5.02 for CRCM6 [33]. The physics packages, including radiation, convection, and boundary layer schemes, have also been updated significantly [34].
Following the LUCAS protocol [19], all simulations presented here were performed at 0.44 ∘ (∼50 km) horizontal resolution with lateral boundary conditions and sea-surface temperature driven by the six-hourly ERA-Interim reanalysis [35]. All three models were run over the North America CORDEX domain, and CRCM5 was run on the Europe CORDEX domain in addition (https://cordex.org (accessed on 11 August 2022)). The simulations are analyzed over 1986–2015, after a seven-year spin-up allowing the models to adjust to land cover modifications. Since the forestation response was not found to change significantly over the analysis period, only climatologies are shown.

2.2. Land Cover

Two simulations were performed for every member of the model-domain ensemble described above: FOREST and GRASS. The only difference between the two is land cover. For both sets of simulations, the vegetation distribution is based on moderate-resolution imaging spectroradiometer (MODIS) land cover maps at 0.5 ∘ resolution [50]. The MODIS maps were then modified according to the protocol outlined in [19] to obtain the FOREST and GRASS land covers. In a nutshell, the FOREST experiment represents the theoretical maximum tree cover: the fractional cover of trees is expanded until trees fill all the area not occupied by bare ground, glaciers, or lakes, which are left untouched. In the process, the proportion of the different tree families is kept fixed. Similarly, all vegetation is replaced by grassland in the GRASS experiment.
The model-dependent land categories used in the conversion are outlined in Table 1. Although the MODIS data is projected onto multiple tree categories in both NOAH and CLASS, two classes dominate (bold in Table 1): deciduous broadleaf trees and evergreen needleleaf trees. Other tree categories appear (italicized in Table 1), but they play a minor role, so we neglect them in the analysis. For the GRASS experiment, all MODIS grasses project onto a single land category. The main properties of these dominant tree and grass categories, which differ between CLASS and NOAH, are outlined in Table 2.
Figure 1 displays the fraction of land occupied by broadleaf and needleleaf trees (left and center panels) in the FOREST experiment, the sum of which gives the grass cover from the GRASS experiment (right panels). The non-grass fraction of the right panels is covered by deserts, glaciers, or lakes. The FOREST versions of North America and Europe reveal a similar pattern: needleleaf forests tend to concentrate at higher latitudes and broadleaf forests at lower latitudes. One difference that will prove important, however, is that these forests appear at lower, hence sunnier, latitudes in North America. In eastern Canada, for instance, latitudes 45 ∘ to 55 ∘ —from the Great Lakes up to Northern Quebec—are densely populated by needleleaf trees in the FOREST world. By comparison, needleleaf trees reach complete coverage only north of 60 ∘ in eastern Europe. Broadleaf forests extend from the Mediterranean region at around 40 ∘ up to 55 ∘ in Europe, while they are mostly concentrated between 30 ∘ and 45 ∘ at most in North America. (We do not include the tropical region in the analysis because it is too close to the domain boundary.)
For subsequent analysis, we define two main forests families: northern needleleaf and southern broadleaf forests. In Figure 1, these forests are delimited by the colored latitudinal and longitude lines. A grid point is considered part of a given forest if it falls between these boundaries and its type accounts for more than 50% of the vegetation fraction. For instance, any grid point of America north of 30 ∘ N, south of 50 ∘ N, and east of 105 ∘ W with more than 50% broadleaf tree cover is considered part of the southern broadleaf forest.

3. Results

3.1. CRCM5 over North America and Europe

We begin with an analysis of the CRCM5-CLASS simulations over North America and Europe. This provides an opportunity for comparing the response to forestation over two different continents, and thus for investigating whether the findings from the original European LUCAS study may apply to North America.

3.1.1. Winter

The CRCM5 winter response to forestation is summarized in Figure 2. In both Europe and North America, forestation causes a widespread wintertime warming, peaking at mid-high latitudes. This warming pattern matches net downwelling shortwave radiation, suggesting that solar energy absorption dominates the temperature signal. Indeed, high-latitude evergreen needleleaf forests are collocated with a strong drop in the shortwave albedo. In CLASS, these trees can intercept snow on their canopy, which increases their albedo. However, snow coverage remains incomplete, and the dark canopy of needleleaf trees can mask snow on the ground. By contrast, grasses may be fully buried by snow. Therefore, snow-covered evergreen forests absorb a much higher fraction of the incoming shortwave radiation than snow-buried grasses, causing warming.
Snow masking was also found to cause wintertime warming in the original LUCAS experiment [19]. Like CRCM5, all LUCAS members simulate a drop in albedo, resulting in warming over the northern evergreen needleleaf forests of Europe (Figure 1 of [19]). We note, however, that CRCM5 has the strongest temperature response of the LUCAS ensemble (see Figure S1). In other words, the wintertime warming mechanism seems robust across models and continents, but one must keep in mind that CRCM5-CLASS may be among the most sensitive to it.
While snow masking is well known and documented in both observational [5,51] and modeling studies [19,52], a more overlooked fact is that its warming effect depends on the strength of insolation, and therefore on the latitude and time of year. Given a fixed reduction of albedo, the effect on net shortwave radiation (and hence warming) will be stronger where and when there is more incoming shortwave. In Figure 2, the albedo drop is uniform over the dense northern needleleaf trees, but the shortwave radiation excess and warming responses decay with increasing latitude. The warming response is strongest at the lowest latitudes where dense evergreen forests are covered by snow. This explains why the shortwave excess and the accompanying warming response are much stronger and longitudinally widespread in North America than in Europe. Dense evergreen needleleaf forests populate broader swaths and lower latitudes in North America (Figure 1). Snow masking by needleleaf trees has the same effect on albedo on both continents, but the resulting shortwave radiation excess and hence warming is much stronger over the sunnier, lower-latitude North American forests.
The left panel of Figure 3 reveals the high correlation—0.85 and 0.75 for North America and Europe, respectively—between warming and insolation over snow-covered needleleaf forests in wintertime. Each dot represents a grid point within the northern needleleaf forests of North America (blue) or Europe (orange) with snow cover during at least 90% of winter. In these regions, the warming response to forestation is dominated by snow masking, and the magnitude of this warming is proportional to incoming shortwave, which is mostly a function of latitude here. Since snow-covered needleleaf forests extend to much lower latitudes in North America, the scatter reaches higher warming and incoming solar input on this continent. The Europe scatter is instead confined to weak warming because needleleaf forests occupy high latitudes with little insolation. It is noteworthy that the two continents have strongly overlapping scatters and trends (0.08 and 0.07 Km2/W for North America and Europe, respectively), suggesting that the mechanisms at work are similar.
Insolation also changes with the time of the year. In springtime, snow cover remains important and incoming shortwave radiation increases both in magnitude and latitudinal reach (Figure S3). Therefore, the warming effect of snow masking is stronger and affects higher latitudes in the springtime for both continents.

3.1.2. Summer

The temperature response is more complex during summer than winter. In both North America and Europe, forestation produces a dipole-like response, with a warming over northern needleleaf forests and a mild cooling over southern broadleaf forests (Figure 4). The response is cooler overall in Europe.
Over northern needleleaf forests, warming regions coincide with a shortwave radiation excess, itself caused by various factors. Firstly, forests—especially needleleaf forests—are darker than grasses. Secondly, the warmer FOREST climate is consistent with precocious snow melt at high latitudes, especially in North America (see Section 3.2). These two effects significantly lower albedo, and thus increase solar energy input at high latitudes. Thirdly, there is a noticeable drop in cloud cover over northern Canada (Figure 5), further increasing incoming solar radiation and hence the warming in this region.
Over the dense lower-latitude broadleaf forests, heavy transpiration instead causes increased cloud cover, which significantly reduces incoming shortwave radiation in central Europe and the eastern US (Figure 5). The cooling spots, however, cannot be fully explained by the weaker, but still positive, net downwelling shortwave radiation. Differences in the partitioning of turbulent heat fluxes play a dominant role in these regions. A useful way to capture these changes is the evaporative fraction (EF):
EF = LH LH + SH
where LH and SH are the latent and sensible heat fluxes—that is, EF represents the fraction of turbulent fluxes due to evapotranspiration.
The right panel of Figure 3 shows the strong negative correlation (−0.79 and −0.9 for North America and Europe, respectively) between the summertime near-surface temperature and EF responses to forestation. To emphasize the effect of land cover changes, the scatter plot comprises all the grid points of the northern needleleaf and southern broadleaf forests defined in Section 2, excluding regions where either the FOREST or the GRASS simulation has over 20 cm of snow on average in the summertime (JJA) period. These isolated snow-covered regions, which appear at or near ground ice sheets, can generate strong EF anomalies that are not related to the land cover changes that we wish to highlight. From Figure 4 and Figure 5, one notes that regions of increasing (decreasing) EF tend to match the location of broadleaf (needleleaf) forests. Mature, unstressed summertime broadleaf trees, with their deeper roots, denser foliage, and weaker stomatal resistance, are more efficient at intercepting rain and pumping water from the soil than needleleaf trees—at least in CLASS (see parameters used in Table 2). As such, broadleaf forests are prone to giving away more of their energy via evapotranspiration than via sensible heat fluxes, leading to a relative cooling of near-surface air. The opposite is true of needleleaf forests: these regions undergo less evaporative cooling, instead giving away more of their energy via sensible heat fluxes, causing near-surface air warming. Satellite-based data also suggest that converting grasses to needleleaf (broadleaf) forests increases sensible (latent) heat fluxes significantly [5].
Changes in EF are related to significant alterations of the summertime water budget (Figure 5; wintertime changes are negligible in comparison). Overall, forestation leads to enhanced evapotranspiration, especially over broadleaf forests. Precipitations are also increased, with comparable contributions from stratiform and convective forms (not shown). In relative terms, FOREST simulations can produce more than a 50% increase in precipitations compared with the GRASS simulations (not shown). Changes are most drastic over Europe, where dense broadleaf forests occupy a broader swath of the continent and deserts are rarer (Figure 1). One also notes that the response patterns of evapotranspiration, precipitations, and cloud coverage tend to be collocated in space.
The CRCM5-LUCAS comparison for summer is challenging. In the original LUCAS experiment of [19], the model ensemble exhibits a wildly divergent summertime temperature response to forestation, including widespread cooling and warming (see Figure 2). Interestingly, the multi-model mean of the LUCAS models is a warming-cooling north-south dipole pattern akin to the CRCM5 response, albeit with weaker amplitude. While this shows that CRCM5 sits well within the LUCAS summertime uncertainty range, it is unclear how one should interpret the mean from such divergent data.
We also note that one of the nine LUCAS members, CCLM-CLM4.5 (short for the COSMO Climate Limited-Area Community RCM coupled to the Community Land Model), responds similarly to CRCM5 in summertime. Both models produce a north-south temperature dipole associated with an inverted EF dipole (see Figure S2). In other words, in both models, southern broadleaf forests favor latent over sensible heat fluxes, and vice versa for northern needleleaf forests. It is unclear, however, what one can learn from this similarity, as two other regional climate models (RegCM and WRF) were coupled to the same land model and yet did not produce a dipole temperature response.
The above remark nevertheless illustrates one of the more robust features across both LUCAS and the present study: the partition of turbulent fluxes plays a major role in determining the summertime temperature response. In Figure 11, from [19], it is shown that a decreased EF is associated with warming (and vice versa) for all models during summertime in Scandinavia. The same is true of CRCM5 on both continents: compare the temperature and EF summertime maps from Figure 4 and Figure 5. While the link between EF and temperature responses is robust, the origin of the inter-model divergence in the partition of turbulent fluxes remains elusive despite considerable efforts [7,8].
These difficulties should not obscure the encouraging implications of the present section. The broad similarity between the CRCM5 responses over North America and Europe suggests that the more robust results from the original LUCAS experiment in Europe may be transferable to North America after correcting for differences in the vegetation distribution (further discussed in Section 4). We also found that the CRCM5 response to forestation sits well within the LUCAS ensemble, with a similar wintertime warming due to snow masking (albeit on the stronger end of the spectrum) and a mixed summertime temperature response largely driven by the partition of turbulent fluxes. This provides confidence regarding the relative skill of CRCM5-CLASS in simulating the main physical processes implicated in the biogeophysical response to forestation.

3.2. Model Intercomparison over North America

One of the main takeaways from earlier model intercomparison studies of land-use change—such as LUCID and LUCAS—is that model intercomparison is crucial indeed. This study does not differ: the response to forestation shows strong inter-model divergence. One cannot pick a single model and hope for an accurate picture of the effects of land-use change. In what follows, all three combinations of regional climate and land surface models (Table 1) are compared over North America. Overall, WRF-NOAH produces a widly different response to the CRCMs, which have relatively similar responses in comparison. Still, despite sharing the same land surface model and parameters, there are nontrivial and interesting differences between CRCM5 and CRCM6.

3.2.1. Winter

Mass afforestation causes widespread winter warming in all models (Figure 6). The CRCMs display a similar warming pattern, reaching peak intensity at mid-high latitudes, consistent with the snow-masking albedo effect of evergreen needleleaf forests (see Section 3.1.1). By comparison, WRF’s response is milder and peaks at lower latitudes.
The shortwave radiation budgets of WRF and CRCMs help explain the source of this temperature difference. WRF shows no shortwave radiation response to forestation at snow-covered high latitudes, whereas the CRCMs may produce upwards of 20 W/m2 excess in these regions. This is because there is essentially no snow-making effect in NOAH at high latitudes: snow fully hides the forest cover as soon as it reaches a depth of 8 cm (or 4 cm for grass). The albedo values of snow are used wherever the snow pack depth is above this threshold, that is, over most of Canada in wintertime. There is thus no albedo difference between the FOREST and GRASS simulations at high-latitudes for WRF.
Instead, the WRF winter warming maximum aligns with an albedo drop in the prairies and US midwest. In this region, the snowpack depth crosses the WRF snow cover threshold, which is higher for the afforested world. This creates a warming feedback whereby the lower albedo of forests increases net shortwave radiation, causing warming and inhibiting the formation of a deep snow pack.

3.2.2. Spring

Forestation causes the strongest warming response during springtime in the CRCMs (Figure 7). Like in winter, the snow-masking albedo effect dominates the signal, but its impact on shortwave radiation excess and hence temperature is supercharged by the much stronger springtime insolation. In Figure 7, all colorbars have been scaled up by 250% compared with the wintertime Figure 6 to avoid complete saturation. Importantly, the magnitude of the albedo drop is similar for spring and winter. However, in spring, the impact on shortwave radiation excess and hence temperature is amplified and extended to much higher latitudes. This echoes a point made in Section 3.1.1: the impact of the snow-masking effect on temperature depends on insolation, which itself depends strongly on latitude and time of the year. The more sunlight there is, the more potent this effect becomes.
The springtime warming response to forestation happens in concert with precocious snow melt. Figure 8 shows the annual cycle of snow line latitude in the FOREST and GRASS simulations. Consistent with LUCAS [21], in CRCMs, forestation causes a large reduction in snow cover during the melting period, but little impact during the accumulation phase.

3.2.3. Summer

Compared with the strong summertime warming-cooling dipole produced by the CRCMs, WRF exhibits a significantly milder and more uniform warming response (Figure 9). This large inter-model divergence in temperature is not, as in winter and spring, mainly captured by the albedo-driven shortwave radiation excess. Summertime turbulent fluxes have a magnitude similar to radiative fluxes, and they also strongly diverge between models (see Figure 10).
Let us begin with what all models agree on. Forests are darker than grasses (Table 2), so albedo is lowered by forestation (Figure 9). The albedo drop is compounded by precocious snow melt over northern Canada in the warmer afforested worlds simulated by the CRCMs. Cloud effects aside, the albedo drop causes an increase in net downwelling shortwave radiation, and thus in the amount of energy available to warm the surface. All models also agree that trees increase the surface roughness (Table 2), thereby facilitating energy transfer back to the atmosphere as turbulent heat fluxes. Indeed, total turbulent fluxes are enhanced almost everywhere by forestation (see Figure 10).
Models diverge strongly, however, in how turbulent heat fluxes are partitioned between their sensible and latent components. Differences in the vegetation parameters (Table 2) can explain some of this divergence, because they influence the evapotranspiration efficiency of the surface. If vegetation is poor at intercepting rain or pumping water from the soil, sensible heat fluxes will likely dominate and transfer heat via convection. In the summertime, the surface is typically warmer than the air above such that sensible heat fluxes warm the atmosphere. If vegetation favors rain interception and re-evaporation, or if it has low canopy resistance and can easily access water in the soil, latent heat fluxes may take over sensible heat fluxes and cause near-surface cooling.
The evapotranspiration efficiency of the dominant plant functional types varies strongly between NOAH and CLASS (Table 2). CLASS broadleaf trees, with the deepest roots, lowest stomatal resistance, and highest roughness length of all vegetation categories, favor transpiration more than needleleaf trees or grasses. They also have the highest maximum leaf area index, making them great at intercepting rain during summer, so more precipitation re-evaporates before reaching the ground. Over eastern US broadleaf forests, CRCMs thus simulate a strong increase of the evaporative fraction, leading to a relative cooling of the near-surface air (Figure 10). In NOAH, needleleaf trees appear to favor evapotranspiration more than both broadleaf trees and grasses. Despite their higher minimum stomatal resistance, they have the deepest roots, and the highest roughness and leaf area index. As such, there is an increase of evaporative fraction and cooling over southeastern US needleleaf forests. In water-stressed regions such as the southwest US, however, a dominance of sensible heat fluxes results in warming (Figure 10).
The partition of turbulent fluxes may also significantly alter the radiative fluxes. Strong latent heat fluxes can stimulate cloud formation, which blocks incoming sunlight, thereby reducing the amount of energy available to warm the surface. This coupling is particularly prominent in CRCM6, in which southern broadleaf forests favor high evapotranspiration rates, generating stronger precipitation and cloud coverage (Figure 11), which drastically reduces incoming shortwave radiation (Figure 9). The net result is a strong cooling of the surface.
The evaporation-precipitation feedback described above shows that one cannot fully explain the partitioning of turbulent fluxes from the parameters of Table 2 alone. CRCM5 and CRCM6 share the same land surface model and parameters, yet produce different turbulent heat fluxes (Figure 10), which in turn feedback on radiative fluxes and temperature (Figure 9). In this particular case, it is plausible that the changes in the parameterizations of the boundary layer and convective processes bear some of the responsibility for the divergence observed (Table 1).

3.2.4. Fall

Forestation has the weakest temperature response during fall (Figure 12). Unlike winter and spring, the snowpack is confined to very high latitudes. Even if snow-masking in CRCMs produces a noticeable albedo drop in Alaska and Northern Canada, the effect on shortwave radiation excess and hence temperature is mild because of weak insolation at those latitudes. Thus, the snow-masking albedo effect that fueled the wintertime and springtime responses is rather weak during fall. Similarly, the dark canopy of northern needleleaf forests does not cause as much warming as in summertime because of the comparatively weaker insolation.
The highest evapotranspiration rate, which made the southern summertime response so dynamic, is also weaker during fall, as deciduous trees lose their foliage. A mild cooling is nevertheless apparent over the eastern US forests. As in summertime, enhanced cloudiness reduces incoming sunlight in CRCMs (not shown).

3.3. Energy Fluxes over Needleleaf and Broadleaf Forests

Most of the maps shown so far reveal patterns resembling the vegetation distribution. In particular, the two main forest families identified earlier in Figure 1—northern needleleaf and southern broadleaf forests— behave in markedly distinct ways. In what follows, the various components of the surface-atmosphere energy fluxes are spatially averaged over these two main forests for both continents, providing a complementary overview of the biogeophysical effects of forestation.

3.3.1. Northern Needleleaf Forests

Evergreen needleleaf forests are darker than grasses (Table 2); hence, they absorb more shortwave radiation. Both the surface and near-surface air warm up in response. This is true for all seasons, models, and continents presented here (Figure 13). Davin et al. [19] produced a similar energy breakdown over Scandinavia (their Figure 9), which loosely fits our northern needleleaf forest region in Europe (Figure 1). They also find forestation to generate excess shortwave radiation and near-surface warming in all models for winter and spring, and for most, but not all models, in summer and fall.
Figure 13 also reveals weak seasonality in how evergreen needleleaf forests spend their excess energy—that is, the breakdown of fluxes for a given model is similar across seasons after scaling for net downwelling shortwave radiation. The one main exception is the WRF-NOAH needleleaf forest, which evacuates most of its shortwave excess energy through latent heat fluxes. Because of this, northern needleleaf forests generate almost no warming (Figure 9 and Figure 13). By contrast, CRCM-CLASS needleleaf trees have a similar response all year round, and favor sensible heat fluxes more than they do latent heat fluxes. As such, these forests spend their excess shortwave energy in a way that causes unabated warming.
While the ratios of the various energy flux components undergo little seasonal variation, the magnitude of the springtime response is outstanding. In Figure 13, the temperature and energy fluxes scales have respectively been scaled up by factors of 2.5 and 5 to avoid overshoot. Compared with winter, springtime insolation is much greater at high latitudes, so the large albedo drop from snow masking produces a stronger shortwave radiation excess and hence warming (see also Figure 7).

3.3.2. Southern Broadleaf Forests

Broadleaf trees are also darker than grasses—albeit less so than needleleaf trees—and thus absorb excess shortwave radiation (Figure 14). The only exception here is for CRCM6 during summer and fall, where incident sunlight is significantly reduced by enhanced cloudiness (Figure 11). This is consistent with the original LUCAS experiment [19], where forestation causes a net shortwave radiation excess in most models and seasons over France and Eastern Europe, the subdomains most similar to our southern broadleaf forest.
Compared with evergreen needleleaf forests, the energy breakdown of deciduous broadleaf forests reveals stronger seasonality. Leafless, dormant wintertime deciduous forests produce lower evapotranspiration rates less than grasses (except in WRF). However, as spring comes, photosynthetically active broadleaf trees transpire and/or intercept rain significantly more than grasses, often causing cooling in the summertime. Summer and fall surface cooling (or weak warming) are associated with the only instances of increasing net downwelling longwave radiation from forestation. That is, the cooler surface gives away less energy via infrared radiation, and/or the cloudier atmosphere radiates more of it back to the surface.

3.3.3. General Remarks

Despite the strong inter-model and seasonal variability in the energy breakdown, a few remarkable patterns emerge. First and foremost, the CRCM5 energy breakdown in Europe is almost always a downscaled version of its North America analog. This is true of both northern needleleaf and southern broadleaf forests. Since both forests populate lower, hence sunnier, latitudes over North America (Figure 1), the primary energy source of the surface is more plentiful there. As a result, forestation invariably causes more warming (or less cooling) over North America. Furthermore, how this additional energy is distributed for a given forest remains more or less unchanged across continents, echoing the similarity between the North America and Europe patterns seen in Section 3.1. This inter-continental consistency in the CRCM5 response to forestation provides an encouraging outlook on the transferability of LUCAS results to North America.
Trees have lower albedo than grasses, an effect compounded by the snow masking of evergreen forests during winter and spring. Cloud effects aside, forestation thus creates an excess of shortwave radiation. This supplemental energy is more easily transferred to the atmosphere via turbulent heat fluxes because of the enhanced roughness of trees. As a result, the sum of sensible and latent heat fluxes is almost always positive in Figure 13 and Figure 14. Except for dormant wintertime leafless deciduous forests, evapotranspiration is stronger over forests than grasslands. While this is consistent with observations [5,53,54,55], the LUCAS ensemble reveals the opposite tendency [19].

4. Discussion

In this paper, we present a new ensemble of regional climate simulations designed to quantify the biogeophysical effects of severe forestation. To do so, we follow the LUCAS protocol [19], whereby climatologies of fully afforested and deforested worlds are compared. Three regional climate models were run over North America, and one of them—CRCM5—was also run over Europe in an attempt to bridge the gap with the original LUCAS experiment [19].
A few robust results emerge that are in line with previous model intercomparison projects, such as LUCAS and LUCID. First, trees being darker than grasses, forestation generally increases the net shortwave radiation input to the surface. Here, the only exception occurs in CRCM6 during summer and fall over the eastern US, as enhanced cloudiness from heavy broadleaf evapotranspiration blocks enough sunlight to cancel the effect of reduced albedo (Figure 9). In winter and spring, the albedo drop is instead compounded by the snow masking of evergreen forests in CRCMs, causing significant warming. A similar albedo-driven warming is seen over northern evergreen forests in the LUCAS experiment, with the CRCM5 producing among the strongest and most widespread of warming responses of the ensemble. We also find forestation to be associated with significant reductions in snow cover during the melting period, as in LUCAS [21].
The importance of turbulent fluxes’ partitioning for the summertime response is also consistent with previous investigations. We find that the ratio of latent heat fluxes to the total turbulent heat fluxes, or evaporative fraction, is inversely related to the surface temperature response (right panel of Figure 3). While we attempt to rationalize the partition seen in our simulations using the basic vegetation parameters such as leaf area index, roughness, albedo, and root depth, a robust understanding of why a given model produces a given partition is cruelly lacking, and remains an outstanding issue [7,8]. Interestingly, even in the case of the two versions of the CRCM, which share the same land model and vegetation parameters, important differences in the turbulent fluxes’ partitioning are seen. We conjecture that these differences may be attributed to updates in the physics parameterizations.
What does this study teach us about the transferability of the original European LUCAS results to North America? Encouragingly, there is strong inter-continent similarity in the CRCM5 response to forestation, which suggests that LUCAS findings may apply to North America after correcting for the differences in the vegetation distributions. One of the main findings in this paper is that both the northern needleleaf and southern broadleaf forests populating Europe and North America appear at lower, hence sunnier, latitudes over the latter. It is thus to be expected that some biogeophysical effects will be magnified, as the primary energy source—sunlight—is more abundant for a given forest family. Inspecting the energy flux breakdowns of Figure 13 (northern needleleaf forests) and Figure 14 (southern broadleaf forests), one finds that the ratio between the different fluxes in CRCM5 is basically the same in Europe and North America, except that the latter is magnified compared to the former. In other words, while forestation implies more extra energy in North America than in Europe, the surface evacuates this extra energy using longwave radiation, sensible and latent heat fluxes in the same proportions on the two continents.
During winter and spring, the masking of snow by evergreen needleleaf trees reduces the surface albedo, generating excess shortwave radiation and hence warming. For a given reduction in albedo, the resulting warming will be proportional to the amount of sunlight received. One thus expects the wintertime warming reported in LUCAS [19] to be amplified over North America. This is what we see with CRCM5: the North America energy breakdown over needleleaf forests is a scaled-up version of its Europe analog (Figure 13). The amount of energy to spend is larger over the sunnier North America, but the way it is distributed across longwave radiation and latent and sensible heat fluxes remains unchanged.
It is important to note, however, that the simulations presented here are forced by reanalyses of the recent past (1986–2015). In a warming world, the snowline will likely retreat to higher latitudes, which would not only limit the area affected by an albedo reduction, but also move this area to less sunny regions of the globe [56]. One thus expects the snow-masking effect of forestation to become less potent over time.
In summertime, the large inter-model divergence both here and in previous intercomparison projects prevents any firm conclusion. We note that broadleaf forests occupy a higher longitudinal fraction of Europe than North America, and that this is associated with a stronger, more widespread cooling spot and precipitation increase over Europe in CRCM5. However, whether this inter-continental difference has robust implications would have to be assessed with a larger ensemble of simulations over North America and Europe.
More generally, it is crucial to recall that this study is based on a very small ensemble of new simulations, that is, three members over North America and only one new member over Europe (although it is complemented by the nine other model combinations from the original LUCAS study). Of the three models used, WRF produced a response that is significantly different from the two versions of the CRCM, whose responses were comparably close. This skewed response distribution with few members, as well as important structural similarities between the CRCM versions, should be kept in mind when assessing the robustness of the results.
Echoing previous intercomparison projects, the present study emphasizes the urgency of constraining and understanding the inter-model divergence in the partition of turbulent heat fluxes. Promising avenues for confronting model output with observations are already being pursued [57]. In spite of these difficulties and uncertainties, this paper shows how substantial biogeophysical effects of severe forestation could be, with up to 50% changes in summertime precipitations and 10 °C springtime warming simulated regionally. If anything, these results remind us that land-based mitigation strategies such as mass forestation cannot only account for carbon sequestration or albedo changes, as is usual today [4].

5. Conclusions

The present study explored the biogeophysical effects of extreme forestation using an ensemble of regional climate models, with a focus on whether the forestation response found in the original European LUCAS experiment [19] might be transferable to North America. In this final section, we briefly outline our main conclusions.
  • The extreme forestation considered in this study has profound biogeophysical effects on the climate of North America and Europe. The response to forestation mainly depends on the season, latitude, vegetation distribution, and the models used to assess it.
  • During winter and spring, forestation causes widespread biogeophysical warming in all models (Figure 6 and Figure 7). Consistent with LUCAS, the warming hotspots occur over snow-masking northern needleleaf forests (except in WRF, which does not represent properly this phenomenon). Snow masking decreases the surface albedo, which increases net shortwave radiation and causes warming.
  • The warming effect of snow masking is proportional to insolation (left panel of Figure 3). Snow-masking reduces albedo uniformly, but its effect on net shortwave excess and warming is magnified at low latitudes and during spring.
  • Since northern needleleaf forests populate lower, hence sunnier, latitudes in North America than in Europe (Figure 1), the snow-masking effect causes more warming there. In other words, a needleleaf tree planted where it grows naturally may cause more biogeophysical warming in North America than in Europe.
  • During the summer and fall, the biogeophysical effects of forestation are more complex and less consistent across models (Figure 9 and Figure 10). The same difficulty appeared in LUCAS and LUCID. The main source of inter-model divergence is the vegetation’s impact on the partitioning of turbulent heat fluxes.
  • Consistent with previous investigations, we find an inverse relationship between the summertime near-surface temperature and evaporative fraction (ratio of latent heat fluxes to total turbulent heat fluxes) responses to forestation (right panel of Figure 3). In other words, cooling tends to occur where forestation promotes evapotranspiration at the expense of sensible heat fluxes.
  • Overall, we find that the main mechanisms driving the biogeophysical response to forestation—snow-masking in winter/spring and changes in the turbulent heat flux partition in summer/fall—are the same in North America and Europe.
  • However, because both forest families (northern needleleaf and southern broadleaf forests) appear at lower latitudes in North America than in Europe, forestation implies more net shortwave radiation excess over North America. The surface ‘spends’ this energy, (that is, distributes it among the other energy fluxes such as longwave radiation and turbulent heat fluxes), in a similar manner over the two continents (Figure 13 and Figure 14).
  • As such, the biophysical effects of extreme forestation are largely transferable between Europe and North America after correcting for differences in the distribution of vegetation.

Supplementary Materials

Reference [19] is cited in the supplementary materials. The following supporting information can be downloaded at: https://www.mdpi.com/article/10.3390/cli1010000/s1, Figure S1: Comparison of the wintertime (DJF) temperature, shortwave, and albedo responses of CRCM5 with some of the other LUCAS models of Davin et al. (2020); Figure S2: Comparison of the summertime (JJA) temperature, shortwave, and evaporative fraction responses of CRCM5 with some of the other LUCAS models of Davin et al. (2020); Figure S3: CRCM5-CLASS springtime response to forestation (FOREST-GRASS) averaged over 1986–2015, MAM. From left to right: near-surface temperature, shortwave radiation excess, surface shortwave albedo, and needleleaf tree distribution; Figure S4: CRCM5-CLASS fall response to forestation (FOREST-GRASS) averaged over 1986–2015, SON. From left to right: near-surface temperature, shortwave radiation excess, surface shortwave albedo, and needleleaf tree distribution.

Author Contributions

Writing—original draft preparation, O.A.; writing—review and editing, O.A., M.L., D.P., A.D.L., M.B. and B.M.; simulation production, O.A., M.G., K.W. and M.B.; Analysis, O.A., M.L., D.P., A.D.L., M.B. and B.M.; project administration, M.L. All authors have read and agreed to the published version of the manuscript.

Funding

This project was made possible thanks to the funding Ouranos received from the Ministère de l’Environnement et de la Lutte contre les changements climatiques (MELCC) in support of the INFO-Crue initiative. This material is based on work supported by NCAR, which is a major facility sponsored by the NSF under Cooperative Agreement No. 1852911. M. Bukovsky’s contributions to this project were funded by the U.S. Department of Energy, Office of Science, Office of Biological and Environmental Research, Regional and Global Climate Modeling Program (grant number DE-SC0016605).

Data Availability Statement

The data and scripts used are available upon request from the corresponding author.

Acknowledgments

OA wishes to thank Nathalie de Noblet-duCoudré for insightful comments and suggestions regarding the manuscript. Computations with the CRCM5 were carried on the SuperMUC-NG supercomputer at the Leibniz Supercomputing Centre (LRZ) of the Bavarian Academy of Sciences and Humanities. CRCM6 computations were carried on the Narval supercomputer hosted at the École de Technologie Supérieure. These computations were enabled in part by the support provided by Calcul Québec (www.calculquebec.ca) and Compute Canada (www.computecanada.ca). NCAR acknowledges high-performance computing support provided by NCAR’s Computational and Information Systems Laboratory (Computational And Information Systems Laboratory, 2017). All simulations are forced using the ERA-Interim reanalysis dataset provided by the Copernicus Climate Change Service (C3S) at the European Centre for Medium-range Weather Forecast (ECMWF), available from https://www.ecmwf.int/en/forecasts/datasets/archive-datasets/reanalysis-datasets/era-interim, accessed on 11 August 2022. We acknowledge the support of LUCAS by WCRP CORDEX as a Flagship Pilot Study, and the cross-domain collaboration within CORDEX FPS LUCAS. We also thank CORDEX FPS LUCAS for providing the simulation protocol and Edouard Davin for providing the FOREST and GRASS vegetation maps.

Conflicts of Interest

The authors declare no conflict of interest.

Abbreviations

The following abbreviations are used in this manuscript:
CLASSCanadian Land Surface Scheme
CORDEXCoordinated Regional Climate Downscaling Experiment
CRCM5Canadian Regional Climate Model, version 5
CRCM6Canadian Regional Climate Model, version 6
EFEvaporative fraction
FPSFlagship pilot study
MODISModerate-resolution imaging spectroradiometer
WRFWeather research and forecast
LUCASLand-Use and Climate Across Scales
LUCIDLand-Use and Climate, Identification of Robust Impacts
RCMRegional climate model
WCRPWorld Climate Research Program

References

  1. Roe, S.; Streck, C.; Obersteiner, M.; Frank, S.; Griscom, B.; Drouet, L.; Fricko, O.; Gusti, M.; Harris, N.; Hasegawa, T.; et al. Contribution of the land sector to a 1.5 C world. Nat. Clim. Chang. 2019, 9, 817–828. [Google Scholar] [CrossRef]
  2. Shukla, P.; Skea, J.; Calvo Buendia, E.; Masson-Delmotte, V.; Pörtner, H.; Roberts, D.; Zhai, P.; Slade, R.; Connors, S.; Van Diemen, R.; et al. IPCC, 2019: Climate Change and Land: An IPCC Special Report on Climate Change, Desertification, Land Degradation, Sustainable Land Management, Food Security, and Greenhouse Gas Fluxes in Terrestrial Ecosystems; Intergovernmental Panel on Climate Change: Geneva, Switzerland, 2019. [Google Scholar]
  3. Perugini, L.; Caporaso, L.; Marconi, S.; Cescatti, A.; Quesada, B.; de Noblet-Ducoudre, N.; House, J.I.; Arneth, A. Biophysical effects on temperature and precipitation due to land cover change. Environ. Res. Lett. 2017, 12, 053002. [Google Scholar] [CrossRef]
  4. Pongratz, J.; Schwingshackl, C.; Bultan, S.; Obermeier, W.; Havermann, F.; Guo, S. Land Use Effects on Climate: Current State, Recent Progress, and Emerging Topics. Curr. Clim. Chang. Rep. 2021, 7, 99–120. [Google Scholar] [CrossRef]
  5. Duveiller, G.; Hooker, J.; Cescatti, A. The mark of vegetation change on Earth’s surface energy balance. Nat. Commun. 2018, 9, 679. [Google Scholar] [CrossRef]
  6. Windisch, M.G.; Davin, E.L.; Seneviratne, S.I. Prioritizing forestation based on biogeochemical and local biogeophysical impacts. Nat. Clim. Chang. 2021, 11, 867–871. [Google Scholar] [CrossRef]
  7. Pitman, A.J.; de Noblet-Ducoudré, N.; Cruz, F.; Davin, E.L.; Bonan, G.; Brovkin, V.; Claussen, M.; Delire, C.; Ganzeveld, L.; Gayler, V.; et al. Uncertainties in climate responses to past land cover change: First results from the LUCID intercomparison study. Geophys. Res. Lett. 2009, 36. [Google Scholar] [CrossRef]
  8. De Noblet-Ducoudré, N.; Boisier, J.P.; Pitman, A.; Bonan, G.; Brovkin, V.; Cruz, F.; Delire, C.; Gayler, V.; Van den Hurk, B.; Lawrence, P.; et al. Determining robust impacts of land-use-induced land cover changes on surface climate over North America and Eurasia: Results from the first set of LUCID experiments. J. Clim. 2012, 25, 3261–3281. [Google Scholar] [CrossRef]
  9. Lejeune, Q.; Seneviratne, S.I.; Davin, E.L. Historical land-cover change impacts on climate: Comparative assessment of LUCID and CMIP5 multimodel experiments. J. Clim. 2017, 30, 1439–1459. [Google Scholar] [CrossRef]
  10. Mearns, L.O.; Arritt, R.; Biner, S.; Bukovsky, M.S.; McGinnis, S.; Sain, S.; Caya, D.; Correia, J.; Flory, D.; Gutowski, W.; et al. The North American regional climate change assessment program: Overview of phase I results. Bull. Am. Meteorol. Soc. 2012, 93, 1337–1362. [Google Scholar] [CrossRef]
  11. Solman, S.A.; Sanchez, E.; Samuelsson, P.; da Rocha, R.P.; Li, L.; Marengo, J.; Pessacg, N.L.; Remedio, A.; Chou, S.; Berbery, H.; et al. Evaluation of an ensemble of regional climate model simulations over South America driven by the ERA-Interim reanalysis: Model performance and uncertainties. Clim. Dyn. 2013, 41, 1139–1157. [Google Scholar] [CrossRef]
  12. Jacob, D.; Petersen, J.; Eggert, B.; Alias, A.; Christensen, O.B.; Bouwer, L.M.; Braun, A.; Colette, A.; Déqué, M.; Georgievski, G.; et al. EURO-CORDEX: New high-resolution climate change projections for European impact research. Reg. Environ. Chang. 2014, 14, 563–578. [Google Scholar] [CrossRef]
  13. Alexandru, A.; Sushama, L. Impact of land-use and land-cover changes on CRCM5 climate projections over North America for the twenty-first century. Clim. Dyn. 2016, 47, 1197–1209. [Google Scholar] [CrossRef]
  14. Mooney, P.; Sobolowski, S.; Lee, H. Designing and evaluating regional climate simulations for high latitude land use land cover change studies. Tellus Dyn. Meteorol. Oceanogr. 2020, 72, 1–17. [Google Scholar] [CrossRef]
  15. Mooney, P.; Lee, H.; Sobolowski, S. Impact of Quasi-Idealized Future Land Cover Scenarios at High Latitudes in Complex Terrain. Earth’s Future 2021, 9, e2020EF001838. [Google Scholar] [CrossRef]
  16. Bukovsky, M.S.; Gao, J.; Mearns, L.O.; O’Neill, B.C. SSP-Based Land-Use Change Scenarios: A Critical Uncertainty in Future Regional Climate Change Projections. Earth’s Future 2021, 9, e2020EF001782. [Google Scholar] [CrossRef]
  17. Li, W.; Li, L.; Chen, J.; Lin, Q.; Chen, H. Impacts of land use and land cover change and reforestation on summer rainfall in the Yangtze River basin. Hydrol. Earth Syst. Sci. 2021, 25, 4531–4548. [Google Scholar] [CrossRef]
  18. Rechid, D.; Davin, E.; de Noblet-Ducoudré, N.; Katragkou, E.; The LUCAS Team. CORDEX Flagship Pilot Study “LUCAS-Land Use & Climate Across Scales”—A new initiative on coordinated regional land use change and climate experiments for Europe. In Proceedings of the 19th EGU General Assembly Conference Abstracts, EGU2017, Vienna, Austria, 23–28 April 2017; p. 13172. [Google Scholar]
  19. Davin, E.L.; Rechid, D.; Breil, M.; Cardoso, R.M.; Coppola, E.; Hoffmann, P.; Jach, L.L.; Katragkou, E.; de Noblet-Ducoudré, N.; Radtke, K.; et al. Biogeophysical impacts of forestation in Europe: First results from the LUCAS (Land Use and Climate Across Scales) regional climate model intercomparison. Earth Syst. Dyn. 2020, 11, 183–200. [Google Scholar] [CrossRef]
  20. Daloz, A.S.; Schwingshackl, C.; Mooney, P.; Strada, S.; Rechid, D.; Davin, E.L.; Katragkou, E.; de Noblet-Ducoudré, N.; Belda, M.; Halenka, T.; et al. Land-atmosphere interactions in sub-polar and alpine climates in the CORDEX FPS LUCAS models: I. Evaluation of the snow-albedo effect. Cryosphere Discuss. 2021, 16, 2403–2419. [Google Scholar] [CrossRef]
  21. Mooney, P.A.; Rechid, D.; Davin, E.L.; Katragkou, E.; de Noblet-Ducoudré, N.; Breil, M.; Cardoso, R.M.; Daloz, A.S.; Hoffmann, P.; Lima, D.C.; et al. Land-atmosphere interactions in sub-polar and alpine climates in the CORDEX FPS LUCAS models: Part II. The role of changing vegetation. Cryosphere Discuss. 2021, 16, 1383–1397. [Google Scholar] [CrossRef]
  22. Breil, M.; Rechid, D.; Davin, E.; de Noblet-Ducoudré, N.; Katragkou, E.; Cardoso, R.; Hoffmann, P.; Jach, L.; Soares, P.; Sofiadis, G.; et al. The opposing effects of reforestation and afforestation on the diurnal temperature cycle at the surface and in the lowest atmospheric model level in the European summer. J. Clim. 2020, 33, 9159–9179. [Google Scholar] [CrossRef]
  23. Sofiadis, G.; Katragkou, E.; Davin, E.L.; Rechid, D.; de Noblet-Ducoudre, N.; Breil, M.; Cardoso, R.M.; Hoffmann, P.; Jach, L.; Meier, R.; et al. Afforestation impact on soil temperature in regional climate model simulations over Europe. Geosci. Model Dev. Discuss. 2021, 15, 1–35. [Google Scholar] [CrossRef]
  24. Breil, M.; Davin, E.L.; Rechid, D. What determines the sign of the evapotranspiration response to afforestation in European summer? Biogeosciences 2021, 18, 1499–1510. [Google Scholar] [CrossRef]
  25. Jach, L.; Warrach-Sagi, K.; Ingwersen, J.; Kaas, E.; Wulfmeyer, V. Land Cover Impacts on Land-Atmosphere Coupling Strength in Climate Simulations With WRF Over Europe. J. Geophys. Res. Atmos. 2020, 125, e2019JD031989. [Google Scholar] [CrossRef]
  26. Jach, L.; Schwitalla, T.; Branch, O.; Warrach-Sagi, K.; Wũlfmeyer, V. Sensitivity of land–atmosphere coupling strength to changing atmosphe ric temperature and moisture over Europe. Earth Syst. Dyn. 2022, 13, 109–132. [Google Scholar] [CrossRef]
  27. Martynov, A.; Laprise, R.; Sushama, L.; Winger, K.; Šeparović, L.; Dugas, B. Reanalysis-driven climate simulation over CORDEX North America domain using the Canadian Regional Climate Model, version 5: Model performance evaluation. Clim. Dyn. 2013, 41, 2973–3005. [Google Scholar] [CrossRef]
  28. Šeparović, L.; Alexandru, A.; Laprise, R.; Martynov, A.; Sushama, L.; Winger, K.; Tete, K.; Valin, M. Present climate and climate change over North America as simulated by the fifth-generation Canadian regional climate model. Clim. Dyn. 2013, 41, 3167–3201. [Google Scholar] [CrossRef]
  29. Skamarock, W.C.; Klemp, J.B.; Dudhia, J.; Gill, D.O.; Barker, D.M.; Wang, W.; Powers, J.G. A Description of the Advanced Research WRF, version 2; Technical Report; National Center For Atmospheric Research Boulder Co Mesoscale and Microscale Meteorology DIV: Boulder, CO, USA, 2005. [Google Scholar]
  30. Verseghy, D.L. CLASS—A Canadian land surface scheme for GCMs. I. Soil model. Int. J. Climatol. 1991, 11, 111–133. [Google Scholar] [CrossRef]
  31. Verseghy, D.; McFarlane, N.; Lazare, M. CLASS—A Canadian land surface scheme for GCMs, II. Vegetation model and coupled runs. Int. J. Climatol. 1993, 13, 347–370. [Google Scholar] [CrossRef]
  32. Côté, J.; Gravel, S.; Méthot, A.; Patoine, A.; Roch, M.; Staniforth, A. The operational CMC–MRB global environmental multiscale (GEM) model. Part I: Design considerations and formulation. Mon. Weather. Rev. 1998, 126, 1373–1395. [Google Scholar] [CrossRef]
  33. Girard, C.; Plante, A.; Desgagné, M.; McTaggart-Cowan, R.; Côté, J.; Charron, M.; Gravel, S.; Lee, V.; Patoine, A.; Qaddouri, A.; et al. Staggered vertical discretization of the Canadian Environmental Multiscale (GEM) model using a coordinate of the log-hydrostatic-pressure type. Mon. Weather. Rev. 2014, 142, 1183–1196. [Google Scholar] [CrossRef]
  34. McTaggart-Cowan, R.; Vaillancourt, P.; Zadra, A.; Chamberland, S.; Charron, M.; Corvec, S.; Milbrandt, J.; Paquin-Ricard, D.; Patoine, A.; Roch, M.; et al. Modernization of atmospheric physics parameterization in Canadian NWP. J. Adv. Model. Earth Syst. 2019, 11, 3593–3635. [Google Scholar] [CrossRef]
  35. Dee, D.P.; Uppala, S.M.; Simmons, A.J.; Berrisford, P.; Poli, P.; Kobayashi, S.; Andrae, U.; Balmaseda, M.; Balsamo, G.; Bauer, D.P.; et al. The ERA-Interim reanalysis: Configuration and performance of the data assimilation system. Q. J. R. Meteorol. Soc. 2011, 137, 553–597. [Google Scholar] [CrossRef]
  36. Janjić, Z.I. The step-mountain eta coordinate model: Further developments of the convection, viscous sublayer, and turbulence closure schemes. Mon. Weather. Rev. 1994, 122, 927–945. [Google Scholar] [CrossRef]
  37. Beljaars, A.; Holtslag, A. Flux parameterization over land surfaces for atmospheric models. J. Appl. Meteorol. Climatol. 1991, 30, 327–341. [Google Scholar] [CrossRef]
  38. Sundqvist, H.; Berge, E.; Kristjánsson, J.E. Condensation and cloud parameterization studies with a mesoscale numerical weather prediction model. Mon. Weather. Rev. 1989, 117, 1641–1657. [Google Scholar] [CrossRef]
  39. Delage, Y. Parameterising sub-grid scale vertical transport in atmospheric models under statically stable conditions. Bound.-Layer Meteorol. 1997, 82, 23–48. [Google Scholar] [CrossRef]
  40. Blackadar, A.K. The vertical distribution of wind and turbulent exchange in a neutral atmosphere. J. Geophys. Res. 1962, 67, 3095–3102. [Google Scholar] [CrossRef]
  41. Mlawer, E.J.; Taubman, S.J.; Brown, P.D.; Iacono, M.J.; Clough, S.A. Radiative transfer for inhomogeneous atmospheres: RRTM, a validated correlated-k model for the longwave. J. Geophys. Res. Atmos. 1997, 102, 16663–16682. [Google Scholar] [CrossRef]
  42. Chou, M.D.; Suarez, M.J. An efficient Thermal Infrared Radiation Parameterization for Use in General Circulation Models; National Aeronautics and Space Administration, Goddard Space Flight Center: Washington, DC, USA, 1994. [Google Scholar]
  43. Li, J.; Barker, H. A radiation algorithm with correlated-k distribution. Part I: Local thermal equilibrium. J. Atmos. Sci. 2005, 62, 286–309. [Google Scholar] [CrossRef]
  44. Kain, J.S.; Fritsch, J.M. A one-dimensional entraining/detraining plume model and its application in convective parameterization. J. Atmos. Sci. 1990, 47, 2784–2802. [Google Scholar] [CrossRef]
  45. Kain, J.S. The Kain–Fritsch convective parameterization: An update. J. Appl. Meteorol. 2004, 43, 170–181. [Google Scholar] [CrossRef]
  46. Bechtold, P.; Bazile, E.; Guichard, F.; Mascart, P.; Richard, E. A mass-flux convection scheme for regional and global models. Q. J. R. Meteorol. Soc. 2001, 127, 869–886. [Google Scholar] [CrossRef]
  47. Hong, S.Y.; Dudhia, J.; Chen, S.H. A revised approach to ice microphysical processes for the bulk parameterization of clouds and precipitation. Mon. Weather. Rev. 2004, 132, 103–120. [Google Scholar] [CrossRef]
  48. Kuo, H.L. On formation and intensification of tropical cyclones through latent heat release by cumulus convection. J. Atmos. Sci. 1965, 22, 40–63. [Google Scholar] [CrossRef]
  49. Bourgouin, P. A method to determine precipitation types. Weather. Forecast. 2000, 15, 583–592. [Google Scholar] [CrossRef]
  50. Lawrence, P.J.; Chase, T.N. Representing a new MODIS consistent land surface in the Community Land Model (CLM 3.0). J. Geophys. Res. Biogeosci. 2007, 112. [Google Scholar] [CrossRef]
  51. Li, Y.; Zhao, M.; Motesharrei, S.; Mu, Q.; Kalnay, E.; Li, S. Local cooling and warming effects of forests based on satellite observations. Nat. Commun. 2015, 6, 6603. [Google Scholar] [CrossRef]
  52. Betts, R.A. Offset of the potential carbon sink from boreal forestation by decreases in surface albedo. Nature 2000, 408, 187–190. [Google Scholar] [CrossRef]
  53. Chen, L.; Dirmeyer, P.A.; Guo, Z.; Schultz, N.M. Pairing FLUXNET sites to validate model representations of land-use/land-cover change. Hydrol. Earth Syst. Sci. 2018, 22, 111–125. [Google Scholar] [CrossRef]
  54. Meier, R.; Davin, E.L.; Lejeune, Q.; Hauser, M.; Li, Y.; Martens, B.; Schultz, N.M.; Sterling, S.; Thiery, W. Evaluating and improving the Community Land Model’s sensitivity to land cover. Biogeosciences 2018, 15, 4731–4757. [Google Scholar] [CrossRef] [Green Version]
  55. Teuling, A.J. A forest evapotranspiration paradox investigated using lysimeter data. Vadose Zone J. 2018, 17, 1–7. [Google Scholar] [CrossRef]
  56. Diro, G.; Sushama, L. Contribution of snow cover decline to projected warming over North America. Geophys. Res. Lett. 2020, 47, e2019GL084414. [Google Scholar] [CrossRef]
  57. Duveiller, G.; Forzieri, G.; Robertson, E.; Li, W.; Georgievski, G.; Lawrence, P.; Wiltshire, A.; Ciais, P.; Pongratz, J.; Sitch, S.; et al. Biophysics and vegetation cover change: A process-based evaluation framework for confronting land surface models with satellite observations. Earth Syst. Sci. Data 2018, 10, 1265–1279. [Google Scholar] [CrossRef] [Green Version]
Climate 10 00138 g001 550
Figure 1. Vegetation fraction (%) of the plant functional types considered in this experiment. The left and center panels are the tree types present in the FOREST experiment, the sum of which gives the grasses fraction in the GRASS experiment, shown in the right panels. Non-grass cover in the right panels is water, glaciers, or bare soil. The orange longitude and latitudes lines delimit the two main forest families considered: southern broadleaf and northern needleleaf forests. A longitude-latitude coordinate is considered part of the given forest if it falls between these boundaries and its type accounts for more than 50% of the vegetation fraction.
Figure 1. Vegetation fraction (%) of the plant functional types considered in this experiment. The left and center panels are the tree types present in the FOREST experiment, the sum of which gives the grasses fraction in the GRASS experiment, shown in the right panels. Non-grass cover in the right panels is water, glaciers, or bare soil. The orange longitude and latitudes lines delimit the two main forest families considered: southern broadleaf and northern needleleaf forests. A longitude-latitude coordinate is considered part of the given forest if it falls between these boundaries and its type accounts for more than 50% of the vegetation fraction.
Climate 10 00138 g001
Climate 10 00138 g002 550
Figure 2. CRCM5-CLASS wintertime response to forestation (FOREST-GRASS) averaged over 1986–2015, DJF. From left to right: near-surface temperature, shortwave radiation excess, surface shortwave albedo, and needleleaf tree distribution.
Figure 2. CRCM5-CLASS wintertime response to forestation (FOREST-GRASS) averaged over 1986–2015, DJF. From left to right: near-surface temperature, shortwave radiation excess, surface shortwave albedo, and needleleaf tree distribution.
Climate 10 00138 g002
Climate 10 00138 g003 550
Figure 3. Left: Wintertime near-surface temperature response to forestation (CRCM5-CLASS) as a function of the incoming shortwave radiation flux (taken from the FOREST simulation). Only grid points within snow-covered (covered by snow for more than 90% of winter in the FOREST simulation) northern needleleaf forests are included. Right: Summertime near-surface temperature response to forestation (CRCM5-CLASS) as a function of the evaporative fraction response to forestation. Only grid points within northern needleleaf and southern broadleaf forests are included. Regions with more than 20 cm snow depth on average are excluded.
Figure 3. Left: Wintertime near-surface temperature response to forestation (CRCM5-CLASS) as a function of the incoming shortwave radiation flux (taken from the FOREST simulation). Only grid points within snow-covered (covered by snow for more than 90% of winter in the FOREST simulation) northern needleleaf forests are included. Right: Summertime near-surface temperature response to forestation (CRCM5-CLASS) as a function of the evaporative fraction response to forestation. Only grid points within northern needleleaf and southern broadleaf forests are included. Regions with more than 20 cm snow depth on average are excluded.
Climate 10 00138 g003
Climate 10 00138 g004 550
Figure 4. CRCM5-CLASS summertime response to forestation (FOREST-GRASS) averaged over 1986–2015, JJA. From left to right: near-surface temperature, shortwave radiation excess, albedo, and broadleaf tree distribution.
Figure 4. CRCM5-CLASS summertime response to forestation (FOREST-GRASS) averaged over 1986–2015, JJA. From left to right: near-surface temperature, shortwave radiation excess, albedo, and broadleaf tree distribution.
Climate 10 00138 g004
Climate 10 00138 g005 550
Figure 5. Summertime water budget response to forestation for CRCM5-CLASS. From left to right: evaporative fraction (1), evapotranspiration, precipitation changes, and cloud cover.
Figure 5. Summertime water budget response to forestation for CRCM5-CLASS. From left to right: evaporative fraction (1), evapotranspiration, precipitation changes, and cloud cover.
Climate 10 00138 g005
Climate 10 00138 g006 550
Figure 6. Most relevant wintertime (DJF) variables: near-surface temperature, excess shortwave radiation, albedo, and snow depth. All panels show the differences between FOREST and GRASS simulations.
Figure 6. Most relevant wintertime (DJF) variables: near-surface temperature, excess shortwave radiation, albedo, and snow depth. All panels show the differences between FOREST and GRASS simulations.
Climate 10 00138 g006
Climate 10 00138 g007 550
Figure 7. Same as Figure 6 but for spring (MAM): near-surface temperature, excess shortwave radiation, albedo, and snow depth. Compared with Figure 6, all colorbars have been scaled up by a factor of 2.5 to avoid complete saturation. All panels show the differences between FOREST and GRASS simulations.
Figure 7. Same as Figure 6 but for spring (MAM): near-surface temperature, excess shortwave radiation, albedo, and snow depth. Compared with Figure 6, all colorbars have been scaled up by a factor of 2.5 to avoid complete saturation. All panels show the differences between FOREST and GRASS simulations.
Climate 10 00138 g007
Climate 10 00138 g008 550
Figure 8. Latitude of the snowline as a function of the month, averaged over 1986–2015. The snowline latitude is defined as the lowest latitude for which over 50% of the land tiles are covered by at least 10 cm of snow. Greenland was removed from this calculation, and the snowline latitude was capped at 75 ∘ (dashed line).
Figure 8. Latitude of the snowline as a function of the month, averaged over 1986–2015. The snowline latitude is defined as the lowest latitude for which over 50% of the land tiles are covered by at least 10 cm of snow. Greenland was removed from this calculation, and the snowline latitude was capped at 75 ∘ (dashed line).
Climate 10 00138 g008
Climate 10 00138 g009 550
Figure 9. Same as Figure 6 but for summer (JJA): near-surface temperature, excess shortwave radiation, albedo, and snow depth. All panels show the differences between FOREST and GRASS simulations.
Figure 9. Same as Figure 6 but for summer (JJA): near-surface temperature, excess shortwave radiation, albedo, and snow depth. All panels show the differences between FOREST and GRASS simulations.
Climate 10 00138 g009
Climate 10 00138 g010 550
Figure 10. Summertime turbulent heat fluxes: evaporative fraction (defined in Equation (1)), latent heat, sensible heat, and their sum. All fluxes are defined as positive when pointing upwards.
Figure 10. Summertime turbulent heat fluxes: evaporative fraction (defined in Equation (1)), latent heat, sensible heat, and their sum. All fluxes are defined as positive when pointing upwards.
Climate 10 00138 g010
Climate 10 00138 g011 550
Figure 11. Summertime precipitation and cloudiness. The leftmost panels show the relative change in precipitation (pr) from forestation, namely 100 % × ( pr FOREST − pr GRASS ) / pr GRASS . The other panels show the conventional FOREST minus GRASS absolute changes for total precipitation (convective plus stratiform), convective precipitation, and cloud fraction.
Figure 11. Summertime precipitation and cloudiness. The leftmost panels show the relative change in precipitation (pr) from forestation, namely 100 % × ( pr FOREST − pr GRASS ) / pr GRASS . The other panels show the conventional FOREST minus GRASS absolute changes for total precipitation (convective plus stratiform), convective precipitation, and cloud fraction.
Climate 10 00138 g011
Climate 10 00138 g012 550
Figure 12. Same as Figure 6 but for fall (SON): near-surface temperature, excess shortwave radiation, albedo, and snow depth. All panels show the differences between FOREST and GRASS simulations.
Figure 12. Same as Figure 6 but for fall (SON): near-surface temperature, excess shortwave radiation, albedo, and snow depth. All panels show the differences between FOREST and GRASS simulations.
Climate 10 00138 g012
Climate 10 00138 g013 550
Figure 13. Surface energy fluxes breakdown over northern needleleaf forests (FOREST-GRASS). Note that the scales are blown up by factors of 2.5 and 5 for the springtime energy fluxes and temperature, respectively. Note that ground heat fluxes are not included, and longwave radiation data was not available for WRF runs.
Figure 13. Surface energy fluxes breakdown over northern needleleaf forests (FOREST-GRASS). Note that the scales are blown up by factors of 2.5 and 5 for the springtime energy fluxes and temperature, respectively. Note that ground heat fluxes are not included, and longwave radiation data was not available for WRF runs.
Climate 10 00138 g013
Climate 10 00138 g014 550
Figure 14. Surface energy fluxes breakdown over northern broadleaf forests. Following [19], radiative fluxes point downward, whereas turbulent fluxes point upward. Note that longwave radiation data were not available for WRF runs.
Figure 14. Surface energy fluxes breakdown over northern broadleaf forests. Following [19], radiative fluxes point downward, whereas turbulent fluxes point upward. Note that longwave radiation data were not available for WRF runs.
Climate 10 00138 g014
Table
Table 1. Main properties of the models used in this paper. See Table 1 from [19] for comparison with the original LUCAS ensemble.
Table 1. Main properties of the models used in this paper. See Table 1 from [19] for comparison with the original LUCAS ensemble.
Model NameWRFCRCM5CRCM6
InstitutionNCAROuranosUQAM
Land Surface SchemeUnified NOAHCLASS v3.5cCLASS v3.6
Land cover classes (principal classes used in FOREST or GRASS are in bold, classes used but with a minor impact are italiziced)1: Urban and Built-up Land1: Evergreen Needleleaf Trees1: Evergreen Needleleaf Trees
2: Dryland Cropland and Pasture2: Evergreen Broadleaf Trees2: Evergreen Broadleaf Trees
3: Irrigated Cropland and Pasture3: Deciduous Needleleaf Trees3: Deciduous Needleleaf Trees
4: Mixed Dryland/Irrigated Cropland and Pasture4: Deciduous Broadleaf Trees4: Deciduous Broadleaf Trees
5. Cropland/Grassland Mosaic5: Tropical Broadleaf Trees5: Tropical Broadleaf Trees
6. Cropland/Woodland Mosaic6: Drought Deciduous Trees6: Drought Deciduous Trees
7. Grassland7: Evergreen Broadleaf Shrub7: Evergreen Broadleaf Shrub
8. Shrubland8: Deciduous Shrubs8: Deciduous Shrubs
9. Mixed Shrubland/Grassland9: Thorn Shrubs9: Thorn Shrubs
10. Savanna10: Short Grass & Forbs10: Short Grass & Forbs
11. Deciduous Broadleaf Forest11: Long Grass11: Long Grass
12. Deciduous Needleleaf Forest12: Crops12: Crops
13. Evergreen Broadleaf13: Rice13: Rice
14. Evergreen Needleleaf14: Sugar14: Sugar
15. Mixed Forest15: Maize15: Maize
16. Water Bodies16: Cotton16: Cotton
17. Herbacious Wetland17: Irrigated Crops17: Irrigated Crops
18. Wooden Wetland18: Urban18: Urban
19. Barren or Sparlsely Vegetated19: Tundra19: Tundra
20. Herbaceous Tundra20: Swamp20: Swamp
21. Wooded Tundra21: Desert21: Desert
22. Mixed Tundra22: Mixed Wood Forests22: Mixed Wood Forests
23. Bare Ground Tundra23: Mixed Shrubs23: Mixed Shrubs
24. Snow or Ice
Conversion method to implement the vegetation maps (FOREST and GRASS)bare soil = 19bare soil = 21bare soil = 21
Needleleaf Evergreen Temperate = 14Needleleaf Evergreen Temperate = 1Needleleaf Evergreen Temperate = 1
Needleleaf Evergreen Boreal = 14Needleleaf Evergreen Boreal = 1Needleleaf Evergreen Boreal = 1
Needleleaf Deciduous Boreal = 12Needleleaf Deciduous Boreal = 3Needleleaf Deciduous Boreal = 3
Broadleaf Evergreen Tropical = 13Broadleaf Evergreen Tropical = 2Broadleaf Evergreen Tropical = 2
Broadleaf Evergreen Temperate = 13Broadleaf Evergreen Temperate = 2Broadleaf Evergreen Temperate = 2
Broadleaf Deciduous Tropical = 11Broadleaf Deciduous Tropical = 4Broadleaf Deciduous Tropical = 4
Broadleaf Deciduous Temperate = 11Broadleaf Deciduous Temperate = 4Broadleaf Deciduous Temperate = 4
Broadleaf Deciduous Boreal = 11Broadleaf Deciduous Boreal = 4Broadleaf Deciduous Boreal = 4
C3 Arctic Grass = 7C3 Arctic Grass = 10C3 Arctic Grass = 10
C3 Grass = 7C3 Grass = 10C3 Grass = 10
C4 Grass = 7C4 Grass = 10C4 Grass = 10
Representation of sub-grid scale vegetation heterogeneityThe dominant class sets the surface properties, other class properties ignoredThe surface properties are averaged over the different classes present on a given tileThe surface properties are averaged over the different classes present on a given tile
Leaf area indexEstimated using seasonally varying green vegetation coverage fraction (FVEG) and the minimum and maximum values for LAI (LAIMIN and LAIMAX, respectively) prescribed for each vegetation type. LAI = ((1.0 − FVEG) * LAIMIN) + (FVEG * LAIMAX)Seasonal cycle with the onset of spring budburst and fall senescence triggered by near-zero values of the air temperature and the first soil layer temperature.Seasonal cycle with the onset of spring budburst and fall senescence triggered by near-zero values of the air temperature and the first soil layer temperature.
Total soil depth and number of hydrologically/thermally active soil layers4 thermally and hydrologically active soil layers with a maximum depth of 2 m17 thermally and hydrologically active soil layers with maximum depth of 5 m.17 thermally and hydrologically active soil layers with maximum depth of 5 m.
Initialization and spin-upInitialization with ERA-InterimInitialization with ERA-InterimInitialization with ERA-Interim
Lateral boundary formulationlinear relaxation10 semi-lag departure points10 semi-lag departure points
Buffer (no. of grid cells)520 grid cells lateral sponge zonegrid cells lateral sponge zone: 10 in longitude, 15 in latitude
No. of vertical levels285671
Turbulence and planetary boundary layer scheme[36]1.5-order closure based on prognostic turbulence kinetic energy [37]. Mixing length based on [38]. Stability functions from [39]1.5-order closure based on prognostic turbulence kinetic energy [37]. Mixing length based on [38], except in laminar conditions where [40] is used. Stability functions from [37].
Radiation schemeLongwave: RRTM [41]; Shortwave: Goddard [42][43][43], with a few updates described in [34]
Deep Convection schemeKain-Fritsch w/ETA trigger [44][45][44], with a few updates described in [34]
Shallow Convection schemen/a[45] transient[46]
Microphysics schemeWSM3 [47][48,49][38,49]
Greenhouse gasesHistoricalHistorical + RCP4.5Historical + RCP8.5
AerosolsPrescribed Observed, UniformPrescribed ObservedHardcoded: higher values ove land than ove ocean; higher values at the equator than at the poles
Table
Table 2. Principal parameters of the main land categories used in CLASS and in NOAH (bold font in Table 1). Minimum and maximum values for leaf area index (LAI) and albedo are tied to the seasonal cycle.
Table 2. Principal parameters of the main land categories used in CLASS and in NOAH (bold font in Table 1). Minimum and maximum values for leaf area index (LAI) and albedo are tied to the seasonal cycle.
CRCM-CLASSNeedleleafBroadleafGrasses
Min. Stomatal Resistance (s/m)250130150
Albedo (visible/near-IR)0.03/0.190.05/0.290.05/0.31
LAI (min–max)1.6–20.5–64
Root Depth (m)121.2
Roughness Length (m)1.520.08
WRF-NOAHNeedleleafBroadleafGrasses
Min. Stomatal Resistance (s/m)12510040
Albedo (min–max)0.120.16–0.170.19–0.23
LAI (min–max)5–6.41.85–3.310.52–2.90
Root Depth (n layers)443
Roughness Length (m)0.50.50.1
Publisher’s Note: MDPI stays neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Share and Cite

MDPI and ACS Style

Asselin, O.; Leduc, M.; Paquin, D.; Di Luca, A.; Winger, K.; Bukovsky, M.; Music, B.; Giguère, M. On the Intercontinental Transferability of Regional Climate Model Response to Severe Forestation. Climate 2022, 10, 138. https://doi.org/10.3390/cli10100138

AMA Style

Asselin O, Leduc M, Paquin D, Di Luca A, Winger K, Bukovsky M, Music B, Giguère M. On the Intercontinental Transferability of Regional Climate Model Response to Severe Forestation. Climate. 2022; 10(10):138. https://doi.org/10.3390/cli10100138

Chicago/Turabian Style

Asselin, Olivier, Martin Leduc, Dominique Paquin, Alejandro Di Luca, Katja Winger, Melissa Bukovsky, Biljana Music, and Michel Giguère. 2022. "On the Intercontinental Transferability of Regional Climate Model Response to Severe Forestation" Climate 10, no. 10: 138. https://doi.org/10.3390/cli10100138

Note that from the first issue of 2016, this journal uses article numbers instead of page numbers. See further details here.

Article Metrics

Back to TopTop