SELES has been applied in a diversity of projects in British Columbia, across Canada, the USA, Finland, Scotland, Spain, Germany, China and elsewhere.

Projects have included theoretical educational, research and practical decision-support tools, ranging from simple models to explore complex systems and emergent properties, to integrated model toolkits to directly support landscape management decisions.

Below is an overview of some projects. See also the publications.

While decision-support is often a stated goal of model development and application, this is often aspirational. I have spent my career building and applying models for decision support, as well as for research and education. Where possible, the projects used for direct decision support include links to a decision outcome (i.e. where modelling had a significant role in the decision process, not vice versa).


Land Use Planning

Spatial forest landscape models built in SELES were used as the primary means to project changes in forest conditions, timber supply and forest-related indicators under multiple potential scenarios, to support several strategic land use planning processes in B.C.

North Coast Land and Resource Management Plan (LRMP) (2001-2006; northwest coastal B.C.). The North Coast Landscape Model was built in collaboration with Don Morgan (B.C. Ministry of Forests) and the B.C. Ministry of Sustainable Resource Management to support the North Coast LRMP process in a one million hectare area in coastal north-western British Columbia, Canada.

Morice Land and Resource Management Plan (2002-2006, 2008, 2021; northern central interior B.C.). The Morice Landscape Model was built in collaboration with the B.C. Ministry of Sustainable Resource Management to support the Morice LRMP process in a one million hectare area in interior north-western British Columbia, Canada.

Haida Gwaii / Queen Charlotte Islands Land Use Plan (2004-2006; 2009-2010; coastal B.C.). The Haida Gwaii Landscape Model was built in collaboration with the B.C. Ministry of Sustainable Resource Management to support the Land Use Planning process in Haida Gwaii / Queen Charlotte Islands, north-western coastal British Columbia, Canada.

Great Bear Rainforest (Coastal B.C.) Land Use Plan (2006-2009, 2014-2016, 2021-2023; coastal B.C.). The models built to support the North Coast, Morice and Haida Gwaii land use planning were integrated and generalized to support land-use planning in the 8 million-plus ha Great Bear Rainforest (GBR). The GBR Land Use Objectives were adopted in 2009, and modelling support has been provided so far for two iterations of review to refine the land use objectives, assess updated data, and explore potential revisions.

Nanwakolas Integrated Resource Management Plan (2022-ongoing; Vancouver Island B.C.). This land use planning process is being done collaboratively between Nanwakolas Council (a group of First Nations), Western Forest Products and others. This work is currently in progress.


Spatial Graphs and Connectivity Analysis

Spatial Graphs are a variant of mathematical graph theory. These have been implemented as a suite of models for cross-grain connectivity analysis, and applied in a number of projects. A variation based on these SELES models, called grainscape, was adapted to and implemented in the R language.

Woodland caribou habitat connectivity (2001-2007): Spatial graphs were used for a connectivity analysis of woodland caribou habitat with Micheline Manseau (Parks Canada and University of Manitoba), and Marie-Josée Fortin (University of Toronto), jointly with use of genetic data, to support planning of a new national park in central Manitoba (Manitoba Lowlands).

Old-growth connectivity in Kennedy Flats (2001-2002): Spatial graphs were used for a connectivity analysis of old growth foresth across Kennedy Flats, Pacific Rim National Park, in collaboration with Long Beach Model Forest, in particular Barb Beasley.

Mountain Pine Beetle host connectivity (2006-2012): Spatial graphs were used for a connectivity analysis of Mountain Pine Beetle host forest across Alberta and Saskatchewan, in collaboration with Terry Shore and Bill Riel (Canadian Forest Service), and Provincial entymologists in Alberta and Saskatchewan.

Spruce Bark Beetle host connectivity (2006-2008): Spatial graphs were used for a connectivity analysis of Spruce Bark Beetle host forest across the entire Yukon Territory, in collaboration with Terry Shore and Bill Riel (Canadian Forest Service) and the Yukon Territory.


Wildlife Habitat and Species at Risk

SELES has been used to implement a wide range of habitat models for many species, ranging from simple calculations to complex spatial analyses.

Snag population dynamics (2002-2005): we designed and implemented a matrix population model based on field data to assess changes in snags (standing dead trees) and related ecological consequences over time of different forest management actions in east-central B.C. (e.g. block size patterns). Done in collaboration with Craig Delong (B.C. Ministry of Forests), and Glenn Sutherland (Cortex Consultants Ltd.).

Spotted Owl recovery planning (2003-2007): a modelling framework, designed in collaboration with Cortex Consultants, B.C. Ministry of Forests and B.C. Ministry of Water, Land and Air Protection was applied to develop a suite of models (including habitat/timber supply, population and connectivity) to assist with the development of a recovery plan for critically endangered Spotted Owls in southwestern B.C. to support the B.C. Spotted Owl Recovery Team.

Grizzly Bear risk assesssment (2015): In collaboration with the B.C. Provincial Grizzly Bear team, we developed models to assess grizzly bear habitat and risk over the entire province of B.C. The core security area model was integrated into the Interim Assessment Protocol for Grizzly Bear in British Columbia, part of the B.C. Cumulative Effects Framework.

Climate refugia assessment (2020-2021): In collaboration with Marvin Eng, Don Morgan and B.C. Ministry of Environment, we developed and implemented methods to identify potential climate refugia at macro and micro scales (based in part on climate velocity) and enduring features, and to combine these into conservation priority classes.


Natural Disturbance

SELES has been used extenstively to model wildfire, insect outbreaks (population models, statistical models, connectivity models) and other types of natural disturbance. Models have ranged from simple, educational-oriented models to complex decision-support models.

Landscape-scale Mountain Pine Beetle outbreak modelling (SELES-MPB; 2000-2008): a spatial landscape-scale mountain pine beetle (MPB) population model combined with a spatial forest management model was built in collaboration with Les Safranyik, Terry Shore and Bill Riel (Canadian Forest Service) and multiple people in the B.C. Ministry of Forests to provide decision-support in response to a vast MPB epidemic that was underway in central British Columbia, Canada. This model was applied in five large landscapes in northern and central B.C. (Kamloops Timber Supply Area, Lakes Timber Supply Area, Morice Timber Supply Area, Williams Lake area, Vanderhoof Forest District), and two landscapes in western Alberta (Foothills Model Forest and Kakwa).

Provincial-scale Mountain Pine Beetle outbreak modelling (BCMPB; 2003-2006): a spatial landscape-scale statistical-based MPB outbreak model (based on annual outbreak observations) was built in collaboration with Les Safranyik, Terry Shore and Bill Riel (Canadian Forest Service), Marvin Eng and Peter Hall (B.C. Ministry of Forests), and Josie Hughes. This model was used for decision-support in response to a vast MPB epidemic in central British Columbia, Canada. It was maintained and used by the Province of B.C. over the duration of the large outbreak to provide inputs to many timber supply review processes (e.g. after 8 years).

Time-based Empirical Fire Model (TEF; 2022-ongoing): The TEF model was designed to assess landscape-scale fire hazard in collaboration with Don Morgan and Phil Burton. One objective is to identify areas of a landscape that are more or less prone to burning (burn likelihood). Fire spread in the TEF model is based on field data collected from large fires in B.C. to estimate relative flammability of fuels, topography and daily wind. This model has a short-term focus using existing forest inventory (possibly modified by potential management actions), but is designed to run over large spatial extents (millions of hectares). Even though there are many different wildfire models, there are relatively few that fill this niche (Burn-P3 is another model with a similar focus).


Sustainable Forest Management

Vermillon landscape model (2000-2010 and 2012): built in collaboration with Christian Messier and Dan Kneeshaw (Université de Québec à Montréal), Marie-Josée Fortin (University of Toronto), Stephen Yamasaki and others to explore sustainable forest management options in boreal landscapes of south central Quebéc, Canada. In 2012, the Vermillon landscape model (VLM) was generalized into the Boreal landscape model (BLM).


Spatial Timber Supply Analysis

The SELES Spatial Timber Supply Model (STSM) was designed in collaboration with Forest Analysis and Inventory Branch, B.C. Ministry od Forests (with particular acknowlegement to Tim Bogle, Christine Fletcher and Kelly Izzard). The SELES STSM (or just STSM) has been used extensively across B.C. to support the timber supply review process (it has been used as the forest estate model in many of the most recent timber supply review analyses for timber supply areas in B.C.).

Development and extension of STSM has been done either as research projects to increase capabilities or during applications to address specific issues. This model is the underlying foundation for the models used to support land-use planning as well as cumulative effects assessments.

Below are some projects in which Gowlland Technologies has been involved.

Robson Valley Landscape Model (2000-2003): built in collaboration with the B.C. Ministry of Forests to explore interactions between natural disturbance (fire, Western Hemlock Looper, Mountain Pine Beetle, Two-Cycle Spruce Budworm) and timber supply in the Robson Valley, east-central British Columbia, Canada. This model was the proto-model of the SELES Spatial Timber Supply Model.

STSM v1 (2001-2003): the first version of the STSM was built in collaboration with Forest Analysis and Inventory Branch, B.C. Ministry for Forests.

Natural disturbance and timber supply (starting in 2002 and ongoing): Integrating natural disturbance into timber supply modelling has been an incremental process, starting with research projects with Marvin Eng and Don Morgan (B.C. Ministry of Forests; 2002-2003). During the large MPB epidemic, in collabration with Forest Analysis and Inventory Branch, B.C. Ministry of Forests, we developed methods to use outputs from the BCMPB model as input to a spatial timber supply analsis (2006-2008). Given the increasing risk posed by climate change, a research project was undertaken to develop methods to improve modelling of natural disturbance in timber supply analysis including integration of potential effects of climate change (2015-2018). This was refined and applied in the timber supply analysis done for the Mackenzie Timber Supply Area (2020).

Wildlife habitat and timber supply (2015-2018): improved methods to represent wildlife habitat objectives in timber supply modelling were developed as part of several research projects with Don Morgan and Christine Fletcher (B.C. Ministry of Forests). These methods include support to represent road density limits, stream crossing objectives, marten territory objectives and operating area controls (e.g. for watershed management).

Haida Gwaii 2020 timber supply review.

Ongoing support: capacity building and model support and setup have been undertaken in several Timber Supply Areas in B.C. (Nass, Kispiox, Morice, Bulkley) to support the timber supply review process.

The most recent release of STSM will be made available on this site at Download STSM.


Cumulative Effects Assessments

The SELES STSM is designed as a toolkit of models (one of which is a forest landscape model, also called a forest estate model or timber supply model). Other tools include hydrological flow models, road processing models, and indicator models.

This toolkit has been used for several Cumulative Effects Assessment (CEA) projects that were designed to explore potential future pathways and trends based on interactions between forest growth, timber harvesting, natural disturbance, pipelines and other linear developments, land-use change and climate change. Effects are assessed as a set of indicators for different values of interest (ecological, social, economic).

Iskut-Stikine CEA (2012-2014): The Iskut-Stikine area of northwestern B.C. has increasing development pressure from mining and gas production. This assessment focused on development of models to account for gas pipeline development and roads for mines (in addition to logging and natural disturbance). Indicators focused on water quality.

Upper Morice River CEA (2014-2016): The Upper Morice River area of northwestern B.C. has increasing development pressure from transportation pipelines and logging as well as increasing impacts from natural disturbance. This assessment focused on development of water balance and glacier balance models, and spatial impacts of roads on water. Indicators focused on salmon and grizzly bear.

Skeena-Nass CEA (2016-2021): The Skeena and Nass watersheds cover a large portion of northwestern B.C., and have increasing development pressure from transportation pipelines and other linear developments, and logging as well as increasing impacts from natural disturbance. This assessment further advanced modelling of grizzly bear core security areas, as well as alternate future scenario pathways.


Forest Carbon Projects

A set of tools have been developed in SELES to translate spatial-temporal outputs from the SELES STSM to the input file formats required for the Carbon Budget Model of the Canadian Forest Sector (CBM-CFS3). These have been used in the Coastal Carbon project that includes the Great Bear Rainforest (GBR) and Haida Gwaii, and which arose from carbon benefit sharing agreements between First Nations and the Province of B.C. This project may be the largest forest carbon project in Canada (in terms of geographic scale).

GBR and Haida Gwaii Carbon Project (2009-present): Estimated forest carbon stock changes over time (using CBM-CFS3) based on landscape projections under the GBR and Haida Gwaii land use orders (and expected in the absence of these land use plans). Forest carbon projections are provided annually to Ostrom Climate Solutions Inc.

Technical input to the B.C. Forest Carbon Offset Protocol (BCFCOP) (2011): Provided input as part of a technical team to the development of the BCFCOP.


Hydrological Flow and Water Balance

Two ways of modelling water flow across a landscape include long-term steady state flow (expected area or volume of water at all locations on a landscape) and short-term water flow dynamics based on weather and climate data.

Steady-state hydrological flow (since 2000): Flow models are used to generate contributing area (the area that flows into each raster cell), as well as a range of other spatial attributes (such as a soil moisture index and terrain stability hazard rating, such as in SinMap). Contributing area can be estimated using a digital elevation model, and can help identify or fill gaps in stream class data (which is often important for ecological values, timber harvest costs, etc.). Even though the result is static, the model uses diffusion to account for flow from all cells, downhill to the edge of the landscape.

Water balance modelling (2014-2016): Water balance models are used to estimate the volume of water flowing across each location of a landscape over time. The model implemented in the Skeena-Nass CEA described in the preceding used monthly spatial rainfall estimates and temperatures (either from historical data or projected future climate) and accounted for how water accumulated (e.g. in snow, ice or lakes) or dissipated over time. This model also included a glacier mass balance component to represent expected response of glaciers to changing precipitation and temperature.