Skip to main content
Dryad

Spatial data for creating a thermal inertia index and incorporating it for conservation applications

Cite this dataset

Boynton, Ryan et al. (2022). Spatial data for creating a thermal inertia index and incorporating it for conservation applications [Dataset]. Dryad. https://doi.org/10.5061/dryad.kwh70rz74

Abstract

This repository contains supporting material for a journal article being submitted to one of the journals published by the American Geophysical Union, titled Earth’s Future. The repository contains the following items:

1. README file of what is in the repository including methods associated with the geodatabase

2.  File Geodatabase

1. README file

The files collected here relate to a study being submitted to the American Geophysical Union’s journal, Earth’s Future. The title of the paper being submitted is, “The contribution of Microrefugia to landscape thermal inertia for climate-adaptive conservation and adaptation strategies.”

The study was conducted across 40,250 km2 of complex mountainous terrain in Northern California. The objective of the study was to consider whether it was possible to identify the relative strength of microrefugia systematically in order to provide conservation and climate-adaptation strategies with information that could help with prioritizing actions. We selected an operational scale of 10 ha (25 acres) as a scale that is suitable for various types of landscape planning exercises, and created a hexagon grid for the region. We calculated the mean value for multiple variables and appended them into the hexagons. For thermal inertia, we calculated the mean elevation per hexagon and then its coolest (highest) point using an environmental lapse rate. We also calculated solar energy loading, calculated the mean solar load per hexagon, and calculated its effect on air temperature. We combined these two temperature metrics to identify how much thermal buffering capacity each hexagon contains, as measured by how much warming it could experience before the mean temperature, as determined from a baseline time period, is no longer found anywhere within the hexagon. We tied the mean annual temperature from 1981–2010 to the mean elevation in each hexagon, as well as a temperature from an earlier period, and from several future periods, based on global circulation models.

The study shows how long current (baseline) climate conditions found in each hexagon may persist and shows how the resulting map of landscape thermal inertia can be used when considering natural vegetation types for conservation, identifying which parts of high-priority wildlife corridors have the greatest capacity to retain their current climate conditions, and what the potential for retaining baseline climate conditions is for areas with late-seral forest conditions as represented by forest canopy height.

The methods section below describes the data used in the study to create the data in the geodatabase that is posted here. The Geodatabase itself provides all the data needed to replicate the various results presented in the paper. Further information can be found in Thorne et al. 2020. That report is more extensive than the results in our associated paper, but it contains more information on the calculation of various metrics associated with and was the foundation from which we developed this study. The report is provided here in order to keep all the relevant materials compiled for potential use by others. 

2. File Geodatabase

The geodatabase is provided as a separate file.

Name: ThermalInertiaIndex.gdb

Contents:

  • AllHexagons
    • A feature class containing all 408,948 hexagon grids used in this study
    • Fields within the feature class:

Id

A unique ID for each hexagon

Watershed

Watershed the hexagon falls within

DomWHR

Habitat type (WHR) that had the majority coverage within the hexagon

WHR_Name

Descriptive name of the habitat type

WHR_GroupName

Major vegetation type

CanopyHt_Score

Canopy Height Score ranging from 1 (under 1m) to 5 (over 25m)

CanopyHt_m

Average canopy height within the hexagon (m)

Conn_Score

Connectivity Score ranging from 1 (low) to 5 (high)

dem10m

Average elevation within the hexagon (m)

dem10m_min

Minimum elevation within the hexagon (m)

dem10m_max

Maximum elevation within the hexagon (m)

SRtemp_min

The lowest Solar Radiation load within the hexagon (degree C)

ElevLR_NegEff2

Effect of elevation on air temperature (degree C)

Thermal_Inertia

Hexagon buffering capacity (degree C)

tave_5180

Average temperature 1951-1980

tave_8110

Average temperature 1981-2010

tave_1039mi8

Average temperature 2010-2039 (MIROC-ESM RCP 8.5)

tave_4069mi8

Average temperature 2040-2069 (MIROC-ESM RCP 8.5)

tave_7099mi8

Average temperature 2070-2099 (MIROC-ESM RCP 8.5)

tave_1039cn8

Average temperature 2010-2039 (CNRM-CM5 RCP 8.5)

tave_4069cn8

Average temperature 2040-2069 (CNRM-CM5 RCP 8.5)

tave_7099cn8

Average temperature 2070-2099 (CNRM-CM5 RCP 8.5)

 

  • Connectivity_Scores
    • 90m raster containing all 3 connectivity scores
    • Fields within the raster:

TNC_Conn_Score

Connectivity Score from reclassed TNC/Omniscape

CEHC_Score

Connectivity Score from reclassed California Essential Habitat Connectivity

Combined_Score

Overall Connectivity Score

 

Methods

These methods describe the steps taken to calculate the attribute columns in the associated database. Compilations were done on publicly available data such as digital elevation models, climate data and others. For references to the public base data used, please see references in Table 1.

There are two sections

a.       How we processed material into the hexagon framework

b.      The sequence of steps for each of the analyses presented in the results section of the main report

a.  How we processed material into the hexagon framework

We created a geodatabase of 10 ha hexagons for the region in order to summarize the spatial data in this study into spatial units that are comparable across the region but that also represent an area size that is relevant for site-level plans such as landscape connectivity or forest conservation.

The hexagon geodatabase covers 28,269 km2 in within the 5 watersheds in northern California, and 40,895 km2 in the 5 watersheds plus a 10 km buffer area.

Integrating data into the hexes

Data from a variety of grid scales, including 10, 30, 90, and 270m was added using the ArcGIS sample tool with the Hexagon centroids to sample the 270m resolution data, and the zonal statistics tool within Hexagon boundaries for raster data with smaller grid cell sizes.

 

This study used four types of data (Table 1):

  • Air temperature & topographic – Topographic data was used to calculate microrefugia buffering capacity for each hexagon. Temperature data was used to evaluate the effect of historical and projected future warming on the ability of local sites to retain baseline temperature conditions.
  • Habitats / Dominant Vegetation Types – Habitat data was used to profile the presence and extent of microrefugia by habitat type for the region
  • Landscape Connectivity Models – were used to find microrefugia in areas that are highly ranked for landscape connectivity
  • Forest Structure data – was used to identify where large, late seral trees occupy microrefugia sites.

Microrefugia – Air temperature & topographic

National Elevation Dataset

www.usgs.gov/core-science-systems/ngp/tnm-delivery

Raster - 10m

Solar Radiation Model

Developed at UC Davis for this study from 25m DEM

Raster - 25m

Environmental Lapse Rate Model

Developed at UC Davis for this study from 10m DEM

Raster - 10m

Linking Temperature to Hexagons

Downscaled PRISM Tmax & Tmin – BCM – current & historical

http://climate.calcommons.org/dataset/2014-CA-BCM

 

Raster – 270 m

Downscaled future climate projections MIROC & CNRM RCP8.5

http://climate.calcommons.org/dataset/2014-CA-BCM

 

Raster – 270 m

Habitats / Dominant Vegetation Types

FVEG - CalFire (FRAP)

 https://frap.fire.ca.gov/mapping/gis-data/

Raster - 30m

Vegetation and Climate Refugia

Vegetative Climate Exposure (UCD Modeling)

Raster - 270m

Landscape Connectivity Models

California Essential Connectivity

https://wildlife.ca.gov/Conservation/Planning/Connectivity/CEHC

Polygon

Omniscape Climate Connectivity

https://omniscape.codefornature.org/

90 m

Forest Structure

Canopy Height - SALO Sciences

https://forestobservatory.com/

Raster - 10m

Table 1: Data sources

b.           The sequence of steps for each of the analyses presented in the results section of the main report

Microrefugia – thermal buffering capacity

Thermal buffering capacity combined two metrics that represent potential modifications to the air temperature in each 10-ha hexagon. First, a 10m digital elevation model was used to calculate the variation in air temperature within each hexagon due to variations in elevation, using a standard environmental lapse rate. Second, the influence of solar radiation on air temperature was calculated. These two metrics were combined.

Elevational Effect on Air Temperature   Column: ElevLR_NegEff2

Zonal Statistics was performed on a 10m DEM for each hex. The range of elevation was used with environmental lapse rate to calculate “buffering capacity” within each Hexagon. We used an environmental lapse rate of 0.00649606 C⁰/ meter (International Civil Aviation Organization, 1993) to calculate the range of temperatures within the hexagon. To calculate the effect of elevation on air temperature within each hexagon we used the following equation: (Average Elevation – Maximum Elevation) x 0.00649606

 

Solar Radiation Effect on Air Temperature: – Column: SRtemp_min

We ran the analysis on a 25 m-resolution DEM. We calculated annualized solar radiation via the r.sun model available in GRASS 7.8 (https://grass.osgeo.org/grass70/manuals/r.sun.html) which calculates direct, diffuse, and reflected solar irradiation for a given day, location, topography, and atmospheric conditions. We assumed clear-sky conditions to run this model, and ran the model for 2 days in each month, from which we calculated solar radiation as a yearly total in watt-hours/m2. We converted the output to megajoules as follows.

Convert yearly watt-hours to daily megajoules

We used a regionally calibrated conversion factor to determine air temperature from Watts/M2/time from a study in the Sierra Nevada Mountains (Curtis et al.; Flint et al. 2021), who determined the relationship of solar radiation with air temperature by using snowmelt and air temperature measurements in the field, which were then used to calibrate a solar radiation model for topographically diverse locations (Flint and Childs, 1987). The best fit resulted in an increase of 0.25 C for every 2.5 MJ/day of solar radiation above 7.5 MJ/day (Table below). Zonal Statistics was performed for each hexagon and the minimum value of solar radiation-driven air temperature was retained.

Daily MJ/sec

Degree C

< 7.5

0

7.5 - 10.5

0.25

10.5 - 12.5

0.5

12.5 - 15

0.75

15 - 17.5

1

17.5 - 20

1.25

20 - 22.5

1.5

22.5 - 25

1.75

25 - 27.5

2

27.5 - 30

2.25

> 30

2.5

 

Calculate thermal buffer by hexagon: – Column: Thermal_Inertia

Hexagon buffering capacity was calculated by adding the temperature equivalent of the lowest Solar Radiation load with the coldest point within each hexagon. The baseline temperature at the mean elevation in each hexagon was defined as its mean temperature for the time period 1981–2010. This permits a view of how long every hexagon can retain local temperatures by examining future climate change, and identification of which hexagons have already warmed more than their buffering capacity by comparing the baseline period to other time periods.

 

Average temperature through time: – Columns: tave_5180 through tave_7099mi8

Minimum and Maximum annual average temperature (tave; 270m rasters) were used from the Basin Characterization Model (BCM) outputs that generate monthly and yearly values. Eight 30-year time periods were calculated. Historic was 1951–1980, baseline is 1981–2010, and the futures are 2010–2039, 2040–2069, and 2070–2099 under 2 GCMs using the RCP8.5 emission scenario and climate change projection data generated for California’s 4th climate vulnerability assessment, the MIROC ESM model which is hotter and drier, and the CNRM CM5 model, which projects warmer and wetter conditions on average for California (Data from Thorne et al., 2017). Average monthly temperatures were calculated by averaging the minimum and maximum temperatures and then compiling the 30-year mean values. Zonal Statistics were then performed for each hex, and average temperature for the 8 30-year time periods mentioned above was retained.

 

Three applications of the microsite thermal buffering are presented in the study: vegetation/habitat types; landscape connectivity; and forest conservation. Each of the applications has data in this geodatabase.

  • Vegetation/habitat types: – Columns: DomWHR, WHR_Name, WHR_GroupName
  • Landscape Connectivity: – Column: Conn_Score
  • Forest Conservation: – Columns: CanopyHt_Score, CanopyHt_m

 

Vegetation/habitat types: – Columns: DomWHR, WHR_Name, WHR_GroupName

Dominant Habitat Types

FVEG (https://frap.fire.ca.gov/mapping/gis-data/), a 30m raster of vegetation types uses the California Wildlife Habitat Relationships System (WHR; https://wildlife.ca.gov/Data/CWHR/Wildlife-Habitats) to classify the vegetation. The FVEG raster was converted to a polygon grid and unionized to our Hexagon grid to determine the dominant WHR type. We used FVEG to calculate:

1.       The major vegetation types. A crosswalk table between WHR types and major vegetation types was used to assign each hexagon a major vegetation type using the following rules:

WHR Code

WHR Name

Major Vegetation Type

ADS

Alpine-Dwarf Shrub

High Elevation Forests and Meadows

AGS

Annual Grassland

Grasslands

ASC

Alkali Desert Scrub

Arid Shrublands

ASP

Aspen

High Elevation Forests and Meadows

BAR

Barren

Barren

BBR

Bitterbrush

Arid Shrublands

BOP

Blue Oak-Foothill Pine

Oak Woodlands

BOW

Blue Oak Woodland

Oak Woodlands

COW

Coastal Oak Woodland

Oak Woodlands

CPC

Closed-Cone Pine-Cypress

Chaparral

CRC

Chamise-Redshank Chaparral

Chaparral

CRP

Cropland

Agriculture

CSC

Coastal Scrub

Chaparral

DFR

Douglas Fir

Conifer Forest

DGR

Dryland Grain Crops

Agriculture

DOR

Deciduous Orchard

Agriculture

EOR

Evergreen Orchard

Agriculture

EPN

Eastside Pine

Conifer Forest

EUC

Eucalyptus

Agriculture

FEW

Fresh Emergent Wetland

Mesic

IGR

Irrigated Grain Crops

Agriculture

IRF

Irrigated Row and Field Crops

Agriculture

IRH

Irrigated Hayfield

Agriculture

JPN

Jeffrey Pine

Conifer Forest

JUN

Juniper

Conifer Forest

KMC

Klamath Mixed Conifer

Conifer Forest

LAC

Lacustrine

Other

LPN

Lodgepole Pine

Conifer Forest

LSG

Low Sage

High Elevation Forests and Meadows

MCH

Mixed Chaparral

Chaparral

MCP

Montane Chaparral

Chaparral

MHC

Montane Hardwood-Conifer

Hardwood-Conifer

MHW

Montane Hardwood

Oak Woodlands

MRI

Montane Riparian

Conifer Forest

PAS

Pasture

Agriculture

PGS

Perennial Grassland

Grasslands

PPN

Ponderosa Pine

Conifer Forest

RFR

Red Fir

Conifer Forest

RIC

Rice

Agriculture

RIV

Riverine

Other

SCN

Subalpine Conifer

High Elevation Forests and Meadows

SGB

Sagebrush

Arid Shrublands

SMC

Sierran Mixed Conifer

Conifer Forest

URB

Urban

Urban

VIN

Vineyard

Agriculture

VOW

Valley Oak Woodland

Oak Woodlands

VRI

Valley Foothill Riparian

Oak Woodlands

WFR

White Fir

Conifer Forest

WTM

Wet Meadow

Mesic

 

2.       The thermal buffering capacity of all hexagons of each vegetation type was then rank-ordered, to identify the extents with different levels of thermal buffer potential.

Landscape Connectivity: – Column: Conn_Score

Landscape & Climate Connectivity

There were two datasets that identify landscape connectivity across the region we were modeling:

1.       The Nature Conservancy (TNC): Omniscape

2.       California Essential Habitat Connectivity (CEHC)

The first examines habitat continuity and tracks analog climates through time (TNC) and the second identifies cores and corridors and then ranks corridors according to how easily an animal could move across the terrain (CEHC).

How was each input reclassed

1.       Omniscape/TNC

a.       Received Connectivity raster data from TNC (https://omniscape.codefornature.org/)

It was a 90m raster with 13 categories so we crosswalked the 13 categories to 4 and ranked them 0-3 (3 is high, 1 is low, 0 is limited connectivity).

Connectivity Score

Description

0

Limited regional connectivity potential

1

Intact landscape

2

Climate linkage (HADGEM2-ES) through an intact landscape

2

Climate linkage (CNRM_CM5) through an intact landscape

3

Climate linkage (both climate models) through an intact landscape

1

Multiple present-day linkage options

2

Climate linkage (HADGEM2-ES) among multiple present-day linkage options

2

Climate linkage (CNRM_CM5) among multiple present-day linkage options

3

Climate linkage (both climate models) among multiple present-day linkage options

1

Present-day linkage

1

Climate linkage (HADGEM2-ES) within a present-day linkage

1

Climate linkage (CNRM_CM5) within a present-day linkage

3

Climate linkage (both climate models) within a present-day linkage

2.       CEHC

a.       Obtained the data from the California Department of Fish and Wildlife (CDFW; https://wildlife.ca.gov/Conservation/Planning/Connectivity/CEHC)

We defined General Natural Landscape Blocks as Cores and Essential Connectivity Areas as Corridors. Corridors were split into 2 groups based on permeability (CDFW; ds620_EssentialConnectivityAreas_CaliforniaEssentialHabitatConnectivity; https://wildlife.ca.gov/Data/BIOS).

We then combined the TNC and CEHC datasets to create an overall connectivity ranking between 0-5 (5 is high, 1 is low, 0 is limited connectivity)

   

CEHC

   

Other areas

Corridor - Less Permeable

Corridor - More Permeable

core

TNC

0

0

1

2

2

1 (low)

1

2

3

3

2

2

3

4

4

3 (high)

3

4

5

5

 

Forest Conservation: – Columns: CanopyHt_Score, CanopyHt_m

Forest Structure/Canopy Height Score

Received Fall 2019 Canopy Height Raster from Salo Sciences (https://forestobservatory.com/). Performed zonal statistics to get the average forest canopy height per hexagon. Classified the heights into 5 classes.

 

Canopy Height Score

Height (m)

1

0 to 1

2

1 to 4

3

4 to 15

4

16 to 25

5

> 25

References:

  • Curtis, J. A., Flint, L.E., Flint, A. L., Lundquist, J. D., Hudgens, B., Boydston, E. E., & Young, J. K. (2014) Incorporating Cold-Air Pooling into Downscaled Climate Models Increases Potential Refugia for Snow-Dependent Species within the Sierra Nevada Ecoregion, CA. PLOS ONE, 9: e106984.
  • Flint, L.E., Flint, A.L., & Stern, M.A. (2021) The Basin Characterization Model version 8 – A Regional Water Balance Software Package. U.S. Geological Survey Techniques and Methods, 6–H1, 85 p., https://doi.org/ 10.3133/ tm6H1.
  • Flint, A. & Childs, S. W. (1987) Calculation of solar radiation in mountainous terrain. Agricultural and Forest Meteorology, 40, 233-249.
  • International Civil Aviation Organization. 1993. Manual of the ICAO Standard Atmosphere (extended to 80 kilometers (3rd edition)). Montréal (Quebec) Canada.
  • Thorne, J. H., Boynton, R. M., Flint, L. E., & Flint, A. L. (2015). The magnitude and spatial patterns of historical and future hydrologic change in California's watersheds. Ecosphere, 6(2), ES14-00300. Online https://esajournals.onlinelibrary.wiley.com/doi/10.1890/ES14-00300.1
  • Thorne, J. H., Boynton, R. M., Wayburn, L., & Urban, D. L. (2020). Planning for Species Adaptation and Climate Resilience in California’s Primary Source Headwaters. Pacific Forest Trust. San Francisco, CA. https://escholarship.org/uc/item/6r2801jn#main

Usage notes

Esri's ArcGIS platform, QGIS, GDAL

Funding

California Wildlife Conservation Board, Award: WC-1835JG