8  Frequently Asked Questions

8.1 If the climate model ensemble mean is different than the historical climate, does this mean the climate models are wrong?

The historical climate is one observation of many possible climate states that could have occurred due to the internal variability (weather) of the Earth’s climate system. Differences between the ensemble mean projection and the observed climate are not necessarily a contradiction; like climate vs. weather, the ensemble mean is the expectation and the observed climate is what actually happened.

However, if the observed climate is completely outside the range of the individual model runs, this is a problem. There are two major reasons why this could occur: (1) the downscaled model ensemble is failing to capture some aspect of regional climate dynamics; or (2) the observations themselves are biased due to errors in the station observations or in the gridded interpolation between stations. Both of these potential sources of error are an active area of research by the CCISS team.

8.2 Does the CCISS tool account for forest health factors?

No. The role of forest health factors is explicitly excluded from the CCISS environmental suitability ratings, with the intention that they are a separate layer of information in reforestation decisions. CCISS is one modeling methodology that considers climate change implications for tree species suitability. Ultimately decisions on species selection should be made using multiple lines of evidence, including insects and disease.

8.3 Why does the CCISS model promote or demote certain species?

The species projected suitability ratings are based on environmental suitability assigned to site series within each BGC subzone-variant. The CCISS model selects BGCs that match the future climate projections for the point(s) or BGC of interest. Future species suitability projections are then pulled from the BGCs projected for that point or area. Go to the “BEC futures” tab to see the breakdown of which BGCs are projected.

8.5 What do C4, B2 and D6 refer to in the CCISS Spatial Module? And why does SNR come first?

These are soil nutrient regime (SNR)/ soil moisture regime (SMR) edatopes, indicating poor/subxeric (B2), mesic/medium (C4) and rich/hygric (D6) site types across all BGCs. Providing spatial data for all 35 edatopes would be too computationally feasible, so we chose these three edatopes to represent distinct, common site types. Although it is typical to place SMR first in edatope names (e.g., 4C), we put SNR first because it is much more convenient in computer code.

8.6 What is the uncertainty around the current suitability ratings for the modeled US BEC units?

There is no equivalent, published biogeoclimatic ecosystem classification and mapping for forest ecosystems of the United States. In the CCISS tool, we use a draft classification of biogeoclimatic units for Washington, Idaho, Montana, Oregon, northern California, and northwestern Wyoming to provide potential future climate analogues for British Columbia (details in MacKenzie & Meidinger 2024 - BC Data Catalogue). Site series classifications were developed, and species suitability ratings were assigned by an expert ecologist based on available plot data from the US and site visits.

There is a higher level of uncertainty associated with the US biogeoclimatic classification as the draft site series classification for US BGC units is more generic and broader (i.e. larger edatopic range) than BC, consequently there is less specificity on tree species suitability. Additionally, biogeoclimatic units are at the subzone level, and in some cases are climatically more diverse than BC units which are at the subzone-variant level. Finally, the species ratings have received less expert review and quantitative validation than those in BC, although they are expert derived and site visits have occurred.

Despite these limitations, the inclusion of these draft US BGC units and associated environmental suitability ratings result in much reduced uncertainty in CCISS projections compared to their absence, as the analogs for future climate in much of southern BC are located across the border in the US. .

8.7 Why doesn’t the US BEC match up with BC BEC across the border?

We are currently integrating the updated BEC classification and mapping for BEC13 (coast region – LMH77). With BEC13, we are also updating mapping of US BEC and are working to align biogeoclimatic mapping across the US – Canada border.

8.8 What is the difference between a projection and a prediction?

“Projection”, “prediction”, and “forecast” are related terms with varying usage in ecology and climate science. For the purposes of CCISS we use the following definitions:

  • Forecast—A forecast is an estimate of possible future states of a variable of interest. A forecast is not a “crystal ball” of what will happen, but a quantification of uncertainty in what could happen. Providing forecasts of environmental suitability for tree species is the ultimate goal of CCISS. CCISS forecasts are a synthesis of many estimates of tree species suitability that have been carefully selected to provide a meaningful representation of uncertainty.

  • Projection—A projection is an estimate of the change over time in a variable of interest given a specific set of assumptions. CCISS provides one projection of tree species suitability for each of three simulations of eight global climate models following three atmospheric greenhouse gas concentration scenarios. The 2001-2020 period of these simulations is considered a projection even though it occurs in the past.

  • Prediction—In CCISS, the term prediction is used exclusively as model prediction: the process by which a statistical model estimates a response variable (i.e., biogeoclimatic subzone/variant analog) based on predictor variables (i.e., climate variables). A model prediction can be made for any values of the predictor variables regardless of whether they represent past, future, or hypothetical conditions. Each biogeoclimatic projection for one 20-year time period is a model prediction.

8.9 In the CCISS spatial popup charts, why do the proportions of suitabilities not match the proportions of BGC analogs?

If there was always only one site series for each edatope, then the proportions in the BGC and suitability plots would match. However, some site series overlap on the edatopic grid or share an edatope. In these cases, there will be two suitabilities for each BGC analog. As a result, the BGC and suitability proportions are not necessarily the same. See Documentation > Methods > Edatopic Overlap.

8.10 In the CCISS spatial popup charts, why do the proportions of suitabilities not match the proportions of BGC analogs?

If there was always only one site series for each edatope, then the proportions in the BGC and suitability plots would match. However, some site series overlap on the edatopic grid or share an edatope. In these cases, there will be two suitabilities for each BGC analog. As a result, the BGC and suitability proportions are not necessarily the same. See Documentation > Methods > Edatopic Overlap.

8.11 What does “decrease of 1” in environmental suitability mean in practical terms? How we should I interpret this in situations like ops planning?

A decrease of 1 means that suitability is expected to decline. For operational purposes, the actual change in rating is important. For example where a species is historically rated an E1 and drops to an E2 - this represents a species that was well within its environmental tolerance (high environmental suitability), which is projected to become moderately suitable - i.e. is outside of its core environmental tolerance range, may be more susceptible to climatic extremes and may have lower growth and survival, but is still suitable and could remain a good choice for planting depending on the specific values/goals for that stand. That would be quite different from a species that was an E3 - low environmental suitability - dropping to “unsuitable”, this would represent a species that was at the edge of its environmental suitability, will likely become unsuitable in the future, and would not be a good choice for planting.

It is important to emphasize that CCISS is a “tool” and should be used alongside local knowledge and within the context of desired values/goals to make reforestation decisions. So a decrease of 1, while at a broad scale means suitability is declining, could lead to different management decisions in practice.

8.12 Are the tree species environmental suitability ratings consistent across all species?

Ratings are applied consistently across species (See “Methods” – “Suitability Ratings”). While they are not perfect, they are expert derived and have gone through an expert review process to achieve a strong scientific foundation (See “Methods” – “Expert Review”). Quantitative validation of the environmental suitability ratings is currently underway (See “Resources”- “CCISS Technical Reports” - “Ratings of Tree Species Environmental Suitability”). Note that environmental suitability does not include forest health factors so those may vary from species to species even where the suitability rating is equivalent.

8.13 Do the environmental suitability ratings represent a linear scale?

Species environmental suitability is not rated on a strictly linear scale. See “Methods” – “Suitability Ratings” for the ratings definitions. Furthermore, increases or decreases in environmental suitability ratings do not necessarily scale linearly with changes in relative abundance for a species, but these metrics are generally positively correlated. Depending on the application it can be useful to look at trends in suitability change. For example, in CCISS Spatial we use the relative change in species suitability to demonstrate broad scale trends in a species across the province. See the CCISS Spatial tab on CCISS and choose “Suitability” - “Future (GCMs)” - “Suitability Change”.

In a case study on Douglas-fir, species environmental suitability ratings showed a positive relationship with relative abundance of the species in the BEC plot dataset. In other words, site series with higher environmental suitability ratings also had higher relative % cover of Douglas-fir. However, the strength of this positive relationship varied by biogeoclimatic zones, with higher agreement in zones where Doughlas-fir is a dominant species at the regional scale, (e.g. Interior Douglas-fir (IDF), Coastal Douglas-fir (CDF) zones) and less agreement in zones with more mixed species composition (e.g. Interior Coastal Hemlock (ICH), and Sub Boreal Spruce (SBS) zones). For more details see “Documents”- “CCISS Technical Reports” - “Ratings of Tree Species Environmental Suitability”).