GEOG 560, GIScience I: Introduction to Geographic Information Science
As anthropogenic climate change continues to progress, locations that are somewhat “disconnected” or “buffered” from regional climate influences may provide a “refugia” for impacted plants and animals. This may give them more time to adapt in-situ or provide them with a migration path to a location with a more suitable climate. Climate change is also predicted to alter the cycles and severities of disturbance regimes, such as droughts, fires and insect infestations. My research is focused on investigating how changes in wildfire disturbance regimes impact riparian climate refugia locations in the topographically diverse inland coast mountains of Northern California.
Keywords used: California, climate change, climate refugia, fire refugia, fire severity, GIS analysis, Landsat, remote sensing, riparian refugia, satellite data.
Annotated Bibliography:
Ackerly, David D., Matthew M. Kling, Matthew L. Clark, Prahlad Papper, Meagan F. Oldfather, Alan L. Flint, and Lorraine E. Flint. 2020. “Topoclimates, Refugia, and Biotic Responses to Climate Change.” Frontiers in Ecology and the Environment 18(5): 288–97.
In this study, the researchers use a combination of spatial and statistical tools, and relevant theoretical models, to develop a climate impact and refugia model for the Mediterranean climate type vegetation community present in the Pepperwood preserve, which is located in California’s north-inland Coastal Mountain Range. The study is based on decades of observations showing that species will generally shift their location, depending on local topographic and soil-related landscape features, in relation to the edges of their distributions, and in relation to their habitat requirements for warm versus cool or dry versus moist areas. The authors use the phrase “hydroclimatic compensation model” (HCM) to describe the interactions of local vegetation community dynamics with the regional climate and the local effects of elevation, topography, soils, and physical factors of the landscape. In this case, the authors specifically studied the impact that climate water deficit (CWD) will have on Pepperwood preserve plant community composition, as the climate continues to change throughout the remainder of the 21st century. The authors obtained geographical image data from the US Forest Service’s Forest Inventory and Analysis (FIA) program site and the Consortium of California Herbaria. Regional climate data was derived from the CHELSA 1- km interpolated climate dataset. The researchers then ran a series of statistical tests to evaluate how 12 focal tree species would be impacted by future climate change in the Pepperwood preserve. The results of the study found that all of the species studied demonstrated a significant correlation with their future range and changes in CWD as a result of climate change. This study provides useful information on how GIS and statistical modeling can be used for analyzing changes in soil moisture, and the impacts that this can have on local vegetation.
Blomdahl, Erika M., Crystal A. Kolden, Arjan J.H. Meddens, and James A. Lutz. 2019. “The Importance of Small Fire Refugia in the Central Sierra Nevada, California, USA.” Forest Ecology and Management 432 (August 2018): 1041–52.
In this paper by Blomdahl et al., the authors highlight some of the issues with detecting “small-scale” (<900 m2) fire refugia using remotely-sensed data. Landsat satellite data has a pixel resolution of 30 m by 30 m, which makes it good for landscape-scale burn pattern analysis, but too low for identifying and analyzing smaller patch areas (i.e., “micro-fire-refugia”). This is particularly true for riparian areas (the focus of my study), since most low order streams have relatively narrow riparian boarder strips that are less than 30 meters across. According to the paper, a common approach used to identify fire disturbance severity is to analyze before and after imagery for net canopy cover changes. Pixels with an unchanged surface reflectance are considered unburned, even if below-canopy understory or groundcover burning occurred in these locations. Conversely, small (<900 m2) unburned patches surrounded by areas of significant canopy cover loss, may be identified as completely burned. In addition, depending on how much time elapsed between before and after imagery, vegetation regrowth can obscure the results. Here, the investigators conducted on-the-ground post-fire surveys in a 25.6 ha plot located within the 104,131 ha California Rim Fire (Aug/Sep 2013), and compared those findings with pre and post fire Landsat satellite imagery for the same location. The study found that Landsat image analysis showed only 0.9% of the study area remained unburned, whereas data from the field observations determined that 4.9% of the area was unburned. In addition, the results found no correlation between burn severity and refugia patch size, however, a positive relationship was found between burn severity and patch density. Riparian areas had a maximum low burn severity ~11 m away from the streambed, suggesting that 30 m by 30 m Landsat imagery data would lack sufficient resolution to identify these low or unburned locations. This study highlights one of the limitations that Landsat satellite data has (i.e., resolution), and how “on-the-ground” field surveys may be required to validate the findings from satellite-based analysis.
Dobrowski, Solomon Z., and Sean A. Parks. 2016. “Climate Change Velocity Underestimates Climate Change Exposure in Mountainous Regions.” Nature Communications 7(1): 1–8.
Dobrowski and Parks define climate change velocity as “the direction and rate at which an organism must move to maintain a given climate envelope over time”. The regional or local velocity of climate change is an important variable for conservation managers, since this will likely impact how long a species can survive in an area before its climate envelope shifts outside of its habitable range. This is especially important for sessile organisms and species with long reproductive cycles. Dobrowski and Parks use an ensemble of 15 CMIP5 global circulation models under the representative concentration pathway 8.5 (commonly referred to as the “business as usual” climate scenario), and GIS modeling techniques, to look at the difference in distance between the “climate trajectory” (defined as a straight line between the current habitat for a species and its closest climate analogue), and the “minimal exposure distance” (defined as the route a species must travel to its closest climate analogue that limits its unfavorable climate exposure to the lowest degree possible), in North American mountainous regions. Using this modeling approach, the authors found that in many cases the direct route to a future climatically suitable habitat is not a straight line, particularly in heterogeneous landscapes. For example, in North America the study found that, on average, the minimal exposure distance was 21% longer than the climate trajectory distance, suggesting that organisms living in heterogeneous mountain landscapes may have a harder time migrating to new locations as their climate envelops shift as a result of climate change. These findings have important implications for land conservation managers, since migration routes that appear to be the most direct, may end up being unusable for the species of interest.
Fraser, Robert H., Ian Olthof, Steven V. Kokelj, Trevor C. Lantz, Denis Lacelle, Alexander Brooker, Stephen Wolfe, and Steve Schwarz. 2014. “Detecting Landscape Changes in High Latitude Environments Using Landsat Trend Analysis: 1. Visualization.” Remote Sensing 6 (11): 11533–57.
In this study, the authors use Landsat satellite data to investigate changes in western Canadian landscapes occurring between 1985 and 2011, which resulted from natural disturbance events, such as wildfires, excessive “greening” of the tundra, geomorphic disturbance events (e.g., the draining of lakes), and the thawing of the permafrost. According to the authors, average temperatures in the region have warmed 2.0 – 2.7°C over the last 50 years, which is 4 to 5 times above global average temperature increases. Landscape types in this part of Canada range from high-boreal forests to low-arctic tundra, and many areas face significant environmental risk from climate change or from more local impacts, such as oil and gas development and the associated buildup of infrastructure. Satellite data was collected for the months of July and August and the images were overlaid into red/green/blue composites for analysis. The results of the study found that wildfires represent the largest disturbance type across the landscape, followed by the expansion of woody shrub growth into the Arctic tundra (mostly due to global warming, according to the authors), and then by the catastrophic drainage of lakes due to the melting permafrost and/or ice dams. The authors point out that the ability to see changes in the landscape over time allowed them to identify disturbance events and trends that might not have been noticed otherwise. This study highlights the value of using time-lapsed satellite imagery for landscape-scale change analysis over time.
Frey, Sarah J.K., Adam S. Hadley, Sherri L. Johnson, Mark Schulze, Julia A. Jones, and Matthew G. Betts. 2016. “Spatial Models Reveal the Microclimatic Buffering Capacity of Old-Growth Forests.” Science Advances 2(4).
In this study, the researchers collected detailed temperature, vegetation structure and topography data within the H.J. Andrews Experimental Forest in Oregon, to study how differences in forest type (i.e., old growth vs. secondary plantation growth) affect below canopy microclimate heterogeneity. Here, the authors used 183 temperature collection gauges along with ground-based LiDAR data for determining the landscape’s topography and vegetation structure. They then used a machine learning technique, known as “boosted regression trees” to ascertain differences in microclimates between the two forest stand types (old growth vs. plantation growth). The results of the study found that the mean temperatures in the old growth stands were more stable, and maximum temperatures during the warm summer months were significantly lower, than those in the plantation stand. The study also found that topography accentuated temperature differentials, such that the low heterogeneity of plantation stands had the highest temperatures at high elevations relative to old growth stands at the same location. These findings suggest that heterogeneous old growth forests may provide high conservation value refugia as climate change continues to progress. This information would be of value to land conservation managers, or others interested in conserving old growth forests.
Gillingham, P. K., B. Huntley, W. E. Kunin, and C. D. Thomas. 2012. “The Effect of Spatial Resolution on Projected Responses to Climate Warming.” Diversity and Distributions 18 (10): 990–1000.
This paper by Gillingham and colleagues looks at how the use of statistical downscaling of the relatively course resolution of most climate forecast models can be used to more accurately predict suitable habitat for species at higher risk from climate change. Here, the authors used a recently published “microclimate model” (oddly, not referenced in the paper) to evaluate its accuracy compared to field measurements of the ground beetle species Carabus glabratus in the Lake Vyrnwy reserve in Wales, U.K. According to the authors, the model uses wind speed, air temperature and solar radiation data along with a digital elevation model having a 5 m by 5 m spatial resolution and 1 m vertical accuracy for the area. The researchers calculated slope and aspect values for each 5 m cell in ArcMap and then used low resolution temperature forecast data to interpolate microclimate temperatures throughout the study area. The results of the study found that as the resolution became finer, the temperature ranges increased, leading the authors to conclude that model sensitivities to climate change are strongly dependent on the spatial resolution of analysis and the model selection method. The study also found that this area of Wales would become uninhabitable for C. glabratus with a 4°C rise in average temperatures. These findings have important climate refugia identification implications, since these areas are generally smaller than the spatial resolution of most climate forecast models.
Hannah, Lee, Lorraine Flint, Alexandra D. Syphard, Max A. Moritz, Lauren B. Buckley, and Ian M. McCullough. 2014. “Fine-Grain Modeling of Species’ Response to Climate Change: Holdouts, Stepping-Stones, and Microrefugia.” Trends in Ecology and Evolution 29(7): 390–97.
In this paper, the authors discuss how differences in microhabitat can influence the interactions between species and their environment. Variables such as temperature, precipitation, solar incidence, wind direction, wind speed, and atmospheric humidity can result in complex mosaics of dynamic micro-ecological environments. As a result, small spatiotemporal environmental fluctuations can occur over short time periods (minutes) and/or short distances (< 1 or 2 meters), that impact how organisms living in these environments respond. Because of the small scales at which these environmental conditions occur, most remotely-sensed imagery is unable to capture this level of detail, and climate forecast models are far too granular to model this in their output. The authors argue that using coarse, mechanistic models, to describe the response of organisms to microclimate conditions can bias estimates of biological change to global warming. They suggest that fine-grained (organism specific) biophysical models that are georeferenced with precision measured climate models can help to eliminate this type of bias. The use of GPS tracking data makes this type of research much more feasible than it was previously. The paper also proposes a new set of terminology to differentiate palaeoecological studies of “climate refugia” from future refugia from anthropogenically-forced climate change. For example, the authors suggest using the term “microholdout” rather than “microrefugium”, and offer the term “micro-stepping-stones” as a way to define a series of microholdouts that facilitate the ability of a given species to shift its range. They also argue that the term “climate refugia” implies a return to pre-disturbance climatic conditions from which organisms can repopulate the surrounding landscape, which may not be the case for a long time with anthropogenic climate change. Finally, the paper proposes that future climate models include algorithms for modeling temperature variations with elevation change (using digital elevation models), which will help biologists and land managers better understand how climate change may impact microholdouts and micro-stepping-stones. However, the paper also points out that computational constraints place limits on how detailed climate forecast models can get in terms of space or time for a given species’ range.
Keppel, Gunnar, Kimberly P. Van Niel, Grant W. Wardell-Johnson, Colin J. Yates, Margaret Byrne, Ladislav Mucina, Antonius G.T. Schut, Stephen D. Hopper, and Steven E. Franklin. 2012. “Refugia: Identifying and Understanding Safe Havens for Biodiversity under Climate Change.” Global Ecology and Biogeography 21(4): 393–404.
This paper looks at the definition of “refuge” and “refugia” in the context of contemporary climate change. Previously, most journal articles defined refugia in the context of long-term glacial cycles, and as locations where, under these relatively slow shifts in the Earth’s climate, allowed plants and animal sufficient time to either adapt in-situ, or move locations along with their shifting climate envelopes. The time between glacial minimum and maximum was on the order of many thousands of years, which gave species ample time to move or adapt. However, under present-day climate change, the climate is changing over a period of a few hundred years, and the pace of change is accelerating as more and more greenhouse gases are emitted into the atmosphere. The authors suggest two methods for identifying and describing present-day refugia. The first is to look for locations that provide a refuge in the past, by collecting and using paleo-biological data found in rocks and soils (along with carbon and other dating techniques) that provide clues to the types of organisms that persisted during times of global change (e.g., glacial maximums). The other approach, which the authors suggest may be better suited to identifying potential refugia from anthropogenic climate change, is to analyze the environmental and physical parameters that set the stage for the emergence and persistence of refugia. For example, by using remotely-sensed data, suitable topographical locations and areas with heterogeneous vegetation cover can be potentially identified as climate refugia locations. The study also points out that remote imagery (e.g., aircraft or satellite obtained), can be integrated with other georeferenced data sources, such as digital elevation models and LiDAR. Together, these tools can those interested in identifying current climate refugia locations, and explains methods for how to monitor those locations as the climate continues to shift in the future.
Keeley, Annika T.H., David D. Ackerly, D. Richard Cameron, Nicole E. Heller, Patrick R. Huber, Carrie A. Schloss, James H. Thorne, and Adina M. Merenlender. 2018. “New Concepts, Models, and Assessments of Climate-Wise Connectivity.” Environmental Research Letters 13(7).
This paper by Keeley et al. discusses the results of their literature review on how to identify “climate-wise” connectivity corridors for “focal species” (i.e., the species of investigative interest), as they need to move in response to “pulse disturbance” events (e.g., wildfire, flood, etc.), or as a result of anthropogenic climate change. Obstacles that impeded (or prevent entirely) the ability of species to shift their range in response to climate change are becoming more and more common due to landscape fragmentation and the speed at which climate change is occurring, which is out-pacing the ability of many species’ to move in pace with their shifting climate envelopes. According to the paper, previous habitat connectivity projects focused on connecting disparate locations of existing suitable habitat for one or more species. For example, two areas of forest that were isolate due to road or urban development might be reconnected via a “connectivity corridor”. However, their focus on “climate-wise connectivity” also includes the objective of connecting habitats that may become suitable in the future, as the climate continues to change. The ability of species to shift their range in conjunction with climate change will be critical for biodiversity conservation in the future, as well as today. Another difference noted by the authors is the directional nature of species range shifts as a result of climate change, which is generally expected to follow changes in temperature and precipitation. Several of the papers in the review used simulation models to evaluate the effects of different landscape designs and species combination, and found that larger habitat areas had generally better short-term conservation outcomes. However, concentrated habitats may not provide sufficient climate variation to allow the focal species to shift their range as their climate envelope moves. The models found that by adding in climate-connection corridors it improved the conservation outcomes in these scenarios. These findings are important to my research because they highlight the need to not only identify potential climate refugia, but also the need to identify ways to connect them to other habitat or refugia locations.
Krosby, Meade, David M. Theobald, Robert Norheim, and Brad H. McRae. 2018. “Identifying Riparian Climate Corridors to Inform Climate Adaptation Planning.” PLoS ONE 13(11): 1–18.
In this paper, Krosby and colleagues provide an overview of the results of their novel approach to modeling riparian climate-corridors (using GIS geospatial analytics) as a way to define categories of how well different riparian sections of stream reaches promote range shifts and provide refugia. Rather than follow the standard practice of calculating a fixed buffer area on either side of a stream reach, the authors used a mapping system developed by Theobald et al. (2013), which quantified potential riparian refugia zones based on areas that: “1) span large temperature gradients, 2) have high levels of canopy cover, 3) are relatively wide, 4) have low solar insolation, and 5) exhibit low levels of human modification”. Using ArcGIS, each raster cell was analyzed and given a value for each of the 5 factors. The results from each factor was then added and the final number was assigned to that cell as a riparian refugia quality index. The results showed that riparian areas located in mountainous landscapes generally had the highest values, whereas flat and low lying areas generally had the lowest values. However, when the authors correlated their calculated riparian refugia quality index with current levels of landscape protection, they found that the low-lying, but high quality areas were also the least protected. This suggests that these riparian areas should be evaluated for conservation opportunities (in addition to more mountainous riparian locations), which would be of interest to land conservation managers.
Mui AB, He Y, Weng Q. 2015. An object-based approach to delineate wetlands across landscapes of varied disturbance with high spatial resolution satellite imagery. ISPRS Journal of Photogrammetry and Remote Sensing 109:30-46
In this study, Mui and team were asked to develop a method for identifying small wetland areas (<0.4 ha) in areas of central Canada. The challenge they faced, is that Landsat data requires (according to the study) a minimum of “9 pure pixels” to accurately and consistently identify a feature. Nine pixels represents approximately 0.9 ha, assuming a 30 m resolution, which results in many wetland areas being missed. A major issue with high-resolution data is the “within-class spectral variance”, which can make it challenging to separate mixed land cover classes, as is the case with coarser-resolution imagery. In this study, the authors used a geographic object based image analysis (GEOBIA) approach/process to convert the pixels into discrete objects. Pixels were classified using a “supervised, nearest neighbor method”, which minimizes the “noise” experienced by high spatial resolution images. The study used this process to analyze three types of landscapes: natural, agricultural and urban. The paper outlines 9 process steps to achieve a final image. These include image preprocessing (radiometric normalization), add supplemental input layers (NDVI, texture and DEM), segment input layers, apply classification attributes, select training samples, classify images, review and refine sample selection (if needed), resegment and assess overall accuracy of the final image. Some steps of the process may be redone depending on the outcome. The results of the processing yielded finer-scale images that resulted in a wetland identification accuracy of more than 80%. This study provides a very nice outline of the step-by-step process the authors used to conduct this analysis, which would be of interest to anyone needing finer resolution data for landscape analysis type projects.
Morelli, Toni Lyn, Sean P. Maher, Marisa C. W. Lim, Christina Kastely, Lindsey M. Eastman, Lorraine E. Flint, Alan L. Flint, Steven R. Beissinger, and Craig Moritz. 2017. “Climate Change Refugia and Habitat Connectivity Promote Species Persistence.” Climate Change Responses 4(1): 1–12.
In this article, the researchers report their findings from a study looking at mountain meadow ecosystems in California’s Sierra Nevada as potential climate refugia for the Belding’s ground squirrel (Urocitellus beldingi). The study also looked at how connectivity between different meadow sites improved persistence of this and other associated species in the face of anthropogenic climate change. The researchers used temperature, precipitation, and snow water equivalent (SWE) PRISM (Parameter-elevation Regressions on Independent Slopes Model) data, which was initially downscaled to 800 m and then additionally downscaled to 270 m using a hydrologic process model. Climate refugia was defined based on a combination of minimal changes in temperatures, precipitation, and SWE from historical averages. Interestingly, this research was based on GIS model output from Maher et al. (2017), which used an application called “Circuitscape” to model potential connectivity corridors based on various “friction surface values”. Model output was then used as the basis for the design of an empirical data gathering process from various locations north and south of Yosemite National Park in California. Field data included site occupancy counts based on visual observation or recognizable Belding’s ground squirrel warning “chirps”. The researchers also collected genetic samples from captured (and subsequently released) squirrels, which was used to evaluate gene flow between populations. The study results suggest that Belding’s ground squirrels are more likely to be found in locations that met the study’s climate change refugia conditions, which may be of interest to conservation managers working in California’s Sierra Nevada.
Maher, Sean P., Toni Lyn Morelli, Michelle Hershey, Alan L. Flint, Lorraine E. Flint, Craig Moritz, and Steven R. Beissinger. 2017. “Erosion of Refugia in the Sierra Nevada Meadows Network with Climate Change.” Ecosphere 8(4).
This paper provides an overview of the process the authors used to develop a geospatial model of interconnected meadow climate refugia in the California Sierra Nevada mountains. The study specifically looked at the connectivity between meadow refugia habitats and assessed the permeability of the surrounding landscape in terms of resistance to, or facilitation of general species dispersal patterns (i.e., the study did not look at any specific species). Connectivity patterns were based on 4 factors that the authors believe impact refugia patch isolation: 1) distance between refugia, 2) topography, 3) water-courses, and 4) roads. Meadow climate refugia were identified using a combination of variables, such as temperature, precipitation and water balance. To identify Sierra Nevada meadows, the authors started with a geodatabase file of 17,037 meadow polygons covering an area of 77,659 ha, across the Sierra Nevada range. The dataset itself was compiled from a list of 44 unique datafiles gathered from a variety of sources by researchers at the University of California, Davis (link to reference if interested). The researches then used ArcGIS to modify the data layer to meet their analysis requirements (e.g., added buffers around meadow sites, upscaled the resolution to improve processing time and reduce variability, etc.). The authors also used an application called “Circuitspace” to develop a connectivity raster layer that uses a concept of electric resistance as a proxy for connection corridor permeability. In addition, the R packages “raster” and “dismo” were used to model the impacts of climate change across the Sierra Nevada, and to identify meadow areas that “qualified” (based on their criteria) as climate refugia. The results of their analysis found a positive correlation between meadow connectivity, meadow elevation and meadow size. The study also found that meadow connectivity progressively increased moving east and north. The use of applications such as Circuitespace for landscape connectivity analysis may be of interest to others doing similar landscape connectivity type projects.
Sommerfeld, Andreas, Cornelius Senf, Brian Buma, Anthony W. D’Amato, Tiphaine Després, Ignacio Díaz-Hormazábal, Shawn Fraver, et al. 2018. “Patterns and Drivers of Recent Disturbances across the Temperate Forest Biome.” Nature Communications 9 (1).
There is still debate among scientists regarding the degree to which forest ecosystem disturbance events are being exacerbated by anthropogenic climate change. For example, were the recent, and unusually large wildfires in Oregon and California driven by the same climate factors as the massive 2019-20 Australian wildfires? Or are the bark beetle infestations occurring in western North American being driven by the same environmental factors as similar infestations in Europe? Sommerfeld and colleagues attempt to answer these types of questions by comparing the severity of disturbance events in protected versus unprotected forested landscapes, as a way to control for anthropogenic versus natural types of disturbances. Forest disturbance imagery data was obtained from the University of Maryland’s Department of Geographical Sciences (http://earthenginepartners.appspot.com/science-2013-global-forest), which utilizes Landsat satellite data to display 30 m2 pixel resolution of changes in forest cover from 2000 to 2014. The researchers then used GIS technology and on-the-ground empirical data to analyze 50 different globally distributed forested regions spanning 16 countries, 5 continents, across both hemispheres, and comprised of a total of 3.9 million hectares. The results found strong evidence to suggest that in protected forest areas with moderate to high severity disturbance events, warmer temperatures result in a significant increase in disturbance probability, and this effect is further amplified in years with below average precipitation. The findings indicate that continued climate change will likely result in more large-scale disturbance events in forest ecosystems as time goes on.
Thorne JH, Boynton RM, Holguin AJ, Stewart JAE, Bjorkman J. 2016. A climate change vulnerability assessment of California’s terrestrial vegetation. California Department of Fish and Wildlife (CDFW), Sacramento, CA.
In this highly detailed report, the authors present their findings from an extensive climate impact risk assessment conducted for California’s 31 natural vegetation community types (referred to as “macrogroups”). For this analysis, Thorne and colleagues evaluated what the impact would be on California’s plant communities under the two most likely climate scenarios: Representative concentration pathway (RCP) 4.5 and RCP8.5, also known as the “business as usual” scenario given it assumes no corrective action is taken to address climate change. To conduct their analysis, the researchers used 30 m vegetation GIS raster files obtained through CalFire (The California State Department of Forestry and Fire Protection) and the Fire and Resource Assessment Program (FRAP). After loading the data into “R”, they ran a series of climate sensitivity analysis for each of the macrogroups. These included analysis of macrogroup sensitivity to temperature increases, precipitation changes (both driven by the climate models), fire sensitivity, dispersal patterns and reproductive lifecycle. The authors then ran the climate model forward and analyzed the shift in each macrogroup’s climate envelope, in order to determine the level of spatial exposure that group had to movement of their respective climate envelops as a result of climate change. They then compiled this data back into a raster data format for visual representation in a GIS. The results of the study found that 93.5% of California’s plant species are a moderate to high risk of climate exposure and spatial disruption of their habitat under the RCP8.5 scenario. This information would be of high value to land conservation managers interested in conserving California’s native vegetation.
Wilkin, Kate M., David D. Ackerly, and Scott L. Stephens. 2016. “Climate Change Refugia, Fire Ecology and Management.” Forests 7 (4).
In this study, Wilkin and colleagues use the ArcGIS spatial autocorrelation analysis tool to identify potential “cold air pools” (CAPs) in Yosemite National Park. Because cold air is denser and heavier than warm air, it flows down into low-lying areas and valley bottoms where it can aggregate and form inversion layers, which prevents warmer air from intermixing. As a result, CAPs generally have lower temperatures and higher moisture levels than the surrounding landscape, and have been identified as potential climate refugia for plants and wildlife occupying these locations. In addition, the authors analyzed spatial data from 1930 to 2012 to identify wildfire burn severity patterns using Relative differenced Normalized Burn Ratio (RdNBR) fire severity categories. High severity wildfires can disrupt the ability of forest canopy to retain cold air and high moisture due to changes in evapotranspiration. After running ArcGIS’s Spatial Autocorrelation, a 100 m point grid was generated using the Create Fishnet geoprocessing tool to analyze the frequency of fire and its severity when it occurs. The data was then loaded into R 3.1.2 for statistical correlation analysis between CAPs and fire frequency and severity. The study’s results found that CAPs were less likely to have been impacted by wildfire as compared to the surrounding landscape, and when they did burn, it was far less destructive than in non-CAP locations. Interestingly, the authors also note that CAPs are frequently found in association with riparian areas, but due to many years of fire suppression activities, these locations may now be a greater risk of high severity wildfires due to the large buildup of surface fuels.
Zellweger F, De Frenne P, Lenoir J, Rocchini D, Coomes D. 2019. Advances in Microclimate Ecology Arising from Remote Sensing. Trends Ecol Evol.
Climate refugia locations can be characterized by the presence of “microclimates”, which Zellweger et al. define as (paraphrased here) areas with a significantly altered climate from the surrounding region due to the influences of local topography and vegetation on the land-air interface. Here, the authors provide a detailed overview of some of the remote sensing technologies and tools used for ecological microclimate analysis. For example, LiDAR (Light Detection and Ranging) is highlighted as a particularly valuable tool for mapping and modeling microclimates due to its high resolution and spatial continuity of vegetation and topography. Similarly, the authors note that terrestrial laser scanning can be used to produce highly detailed datasets of vegetation structure, which can then be used in conjunction with airborne laser scanning to produce microclimate land-cover models. In turn, this data can be used to develop detailed 3D vegetation models that can be analyzed under varying simulated diurnal conditions, different weather conditions, different seasons and under various other climatological conditions. Thermal imaging cameras can also be used to map surface, vegetation, and wildlife temperatures in microclimate locations. Thermal imagery data can also be used to help identify transpiration rates in plants, or to measure the temperature experienced by insects living on leaf surfaces. In addition, thermal imaging has been shown to highlight the potential presence of disease, or to assess the thermal buffering capacity of a forest after selective logging. Finally, macroclimate data can be downscaled to much finer resolutions using high-resolution digital terrain models, canopy elevation models, insolation indexes, cold air pools, and hydrologic conditions.
Additional Resources:
Aaron-Morrison AP, Ackerman SA, Adams NG, Adler RF, Albanil A, Alfaro EJ, Allan R, Alves LM, Amador JA, Andreassen LM, et al. 2017. State of the climate in 2016. Bulletin of the American Meteorological Society. 98(8):Si-S280.
Allen RJ, Anderson RG. 2018. 21st century California drought risk linked to model fidelity of the El Niño teleconnection. Climate and Atmospheric Science. 1(1).
Bertrand, Romain, Gabriela Riofrío-Dillon, Jonathan Lenoir, Jacques Drapier, Patrice De Ruffray, Jean Claude Gégout, and Michel Loreau. 2016. “Ecological Constraints Increase the Climatic Debt in Forests.” Nature Communications 7.
Bowler DE, Mant R, Orr H, Hannah DM, Pullin AS. 2012. What are the effects of wooded riparian zones on stream temperature? Environmental Evidence. 1(1):1–9.
Brown JK, Smith. JK. 2000. Wildland fire in ecosystems: effects of fire on flora. Gen Tech Rep RMRS-GTR-42. 2:257. http://stream.fs.fed.us/news/streamnt/pdf/SN_04_06.pdf.
Capon SJ, Chambers LE, MacNally R, Naiman RJ, Davies P, Marshall N, Pittock J, Reid M, Capon T, Douglas M, et al. 2013. Riparian Ecosystems in the 21st Century: Hotspots for Climate Change Adaptation? Ecosystems. 16(3):359–381.
Coumou, Dim, and Stefan Rahmstorf. 2012. “A Decade of Weather Extremes.” Nature Climate Change 2 (7): 491–96.
Crausbay SD, Martin PH. 2016. Natural disturbance, vegetation patterns and ecological dynamics in tropical montane forests. Journal of Tropical Ecology. 32(5):384–403.
Donat MG, Lowry AL, Alexander L v., O’Gorman PA, Maher N. 2016. More extreme precipitation in the world’s dry and wet regions. Nature Climate Change. 6(5):508–513.
Ellis CJ. 2020. Microclimatic refugia in riparian woodland: A climate change adaptation strategy. Forest Ecology and Management. 462(February).
Fraterrigo JM, Langille AB, Rusak JA. 2020. Stochastic disturbance regimes alter patterns of ecosystem variability and recovery. PLoS ONE. 15(3):1–20.
Godsey SE, Kirchner JW, Tague CL. 2014. Effects of changes in winter snowpacks on summer low flows: Case studies in the Sierra Nevada, California, USA. Hydrological Processes. 28(19):5048–5064.
Goss M, Swain DL, Abatzoglou JT, Sarhadi A, Kolden CA, Williams AP, Diffenbaugh NS. 2020. Climate change is increasing the likelihood of extreme autumn wildfire conditions across California. Environmental Research Letters. 15(9).
Howat IM, Tulaczyk S. 2005. Trends in spring snowpack over a half-century of climate warming in California, USA. Annals of Glaciology. 40:151–156. doi:10.3189/172756405781813816.
Ishida K, Gorguner M, Ercan A, Trinh T, Kavvas ML. 2017. Trend analysis of watershed-scale precipitation over Northern California by means of dynamically-downscaled CMIP5 future climate projections. Science of the Total Environment. 592:12–24.
Keppel G, Mokany K, Wardell-Johnson GW, Phillips BL, Welbergen JA, Reside AE. 2015. The capacity of refugia for conservation planning under climate change. Frontiers in Ecology and the Environment. 13(2):106–112. doi:10.1890/140055.
Keshtkar H, Voigt W. 2016. Potential impacts of climate and landscape fragmentation changes on plant distributions: Coupling multi-temporal satellite imagery with GIS-based cellular automata model. Ecological Informatics. 32:145–155.
Krawchuk MA, Haire SL, Coop J, Parisien MA, Whitman E, Chong G, Miller C. 2016. Topographic and fire weather controls of fire refugia in forested ecosystems of northwestern North America. Ecosphere. 7(12).
Krawchuk MA, Meigs GW, Cartwright JM, Coop JD, Davis R, Holz A, Kolden C, Meddens AJH. 2020. Disturbance refugia within mosaics of forest fire, drought, and insect outbreaks. Frontiers in Ecology and the Environment. 18(5):235–244.
Lazarus ED, McGill BJ. 2014. Pushing the pace of tree species migration. PLoS ONE. 9(8). doi:10.1371/journal.pone.0105380.
Littlefield CE, McRae BH, Michalak JL, Lawler JJ, Carroll C. 2017. Connecting today’s climates to future climate analogs to facilitate movement of species under climate change. Conservation Biology. 31(6):1397–1408.
Luo L, Apps D, Arcand S, Xu H, Pan M, Hoerling M. 2017. Contribution of temperature and precipitation anomalies to the California drought during 2012–2015. Geophysical Research Letters. 44(7):3184–3192.
McDowell NG, Allen CD, Anderson-Teixeira K, Aukema BH, Bond-Lamberty B, Chini L, Clark JS, Dietze M, Grossiord C, Hanbury-Brown A, et al. 2020. Pervasive shifts in forest dynamics in a changing world. Science. 368(6494).
Morelli TL, Barrows CW, Ramirez AR, Cartwright JM, Ackerly DD, Eaves TD, Ebersole JL, Krawchuk MA, Letcher BH, Mahalovich MF, et al. 2020. Climate-change refugia: biodiversity in the slow lane. Frontiers in Ecology and the Environment. 18(5):228–234. doi:10.1002/fee.2189.
Netherer S, Panassiti B, Pennerstorfer J, Matthews B. 2019. Acute Drought Is an Important Driver of Bark Beetle Infestation in Austrian Norway Spruce Stands. Frontiers in Forests and Global Change. 2(July):1–21.
Pierce DW, Kalansky JF, Cayan DR. 2018. Climate, Drought, and Sea Level Rise Scenarios for California’s Fourth Climate Change Assessment. Scripps Institution of Oceanography), California Energy Commission.(August 2018).
Sanchez-Lugo A, Morice C, Nicolas JP, Arguez A. 2020. Global Climate. In: Dunn RJH, Stanitski DM, Gobron N, Willett KM, editors. State of the Climate in 2019. Vol. 101. p. S9–S128.
Thorne JH, Gogol-Prokurat M, Hill S, Walsh D, Boynton RM, Choe H. 2020. Vegetation refugia can inform climate-adaptive land management under global warming. Frontiers in Ecology and the Environment. 18(5):281–287.
Trenberth, Kevin E. 2011. “Changes in Precipitation with Climate Change.” Climate Research 47 (1–2): 123–38.
Trenberth KE, Fasullo JT, Shepherd TG. 2015. Attribution of climate extreme events. Nature Climate Change. 5(8):725–730.
Turner MG, Braziunas KH, Hansen WD, Harvey BJ. 2019. Short-interval severe fire erodes the resilience of subalpine lodgepole pine forests. Proceedings of the National Academy of Sciences of the United States of America. 166(23):11319–11328.
Viles HA, Naylor LA, Carter NEA, D. C. 2008. Biogeomorphological disturbance regimes: progress in linking ecological and geomorphological systems. Earth Surface Processes and Landforms. 33:1419–1435. doi:10.1002/esp.1717.
Wang C, Dong S, Evan AT, Foltz GR, Lee SK. 2012. Multidecadal covariability of North Atlantic sea surface temperature, African dust, Sahel Rainfall, and Atlantic hurricanes. Journal of Climate. 25(15):5404–5415.
Williams AP, Abatzoglou JT, Gershunov A, Guzman-Morales J, Bishop DA, Balch JK, Lettenmaier DP. 2019. Observed Impacts of Anthropogenic Climate Change on Wildfire in California. Earth’s Future. 7(8):892–910.
Zandalinas SI, Mittler R, Balfagón D, Arbona V, Gómez-Cadenas A. 2018. Plant adaptations to the combination of drought and high temperatures. Physiologia Plantarum. 162(1):2–12.