Calgary’s Microclimatic Anatomy: What Lies Beneath the City’s Green Appearance?

Half of Calgary’s area consists of artificial and impervious surfaces: asphalt, roofs, parking lots, and paved areas. If non-residential zones are excluded to focus solely on residential communities, the situation becomes even more striking: the share of non-vegetated areas rises to 63.24%. In other words, an average Calgary residential community is two-thirds sealed by concrete and asphalt.

Dense tree and shrub canopy (Forest) accounts for only 2% of the city’s area. Sparse trees and parkland plantings (Park) make up 15.07%. Thus, the area of grid cells dominated by closed and sparse tree canopy (Forest and Park classes) constitutes ~17% of the total city area, while grassy lawns (Lawn) occupy 28.58%. However, during a dry Calgary summer, lawns quickly dry out and practically cease cooling the air (unlike trees with deep root systems).

In the first part of the Calgary Urban Heat Islands study, spatial land surface temperature (LST) anomalies and spectral vegetation indices were analyzed. However, the NDVI index alone does not provide a complete answer regarding the mechanisms of heating: it does not distinguish, for example, between closed tree canopy and a well-manicured lawn, and it groups dry soil, asphalt, and water into similar categories. To understand why specific Calgary neighborhoods experience significant overheating, it is necessary to examine the physical structure of the underlying surface (Land Use / Land Cover — LULC).

To achieve this, a machine learning model was developed to classify the entire city at a detailed spatial resolution of 10 × 10 meters—a scale that allows assessing the structure and dominant land cover type down to individual residential lots, as well as separating tall tree and shrub vegetation from flat grass cover.

The model operates with five physically interpretable land cover classes:

• Solid (Non-vegetated hard surfaces): asphalt, concrete pads, building roofs, parking lots, bare soil, and construction sites.
• Lawn (Grassy cover): open lawns, turfgrass, low shrubs.
• Park (Sparse tree canopy): large standalone trees, parkland plantings with open canopies.
• Forest (Dense forest): closed tree canopy with dense understory.
• Water (Water bodies): rivers, lakes, reservoirs, stormwater ponds.

Land Cover Across the Entire City Territory

Analysis of the entire municipal area (including peripheral lands, industrial zones, infrastructure facilities, and logistics hubs) shows a dominance of non-vegetated surfaces:

• Solid: 51.94% (more than half of the total city area)
• Lawn: 28.58%
• Park: 15.07%
• Water: 2.41%
• Forest: 2.00%

Total developed tree canopy (Forest + Park) accounts for only 17.07%, whereas more than half of the city’s area is devoid of any vegetation and directly exposed to solar heating.

Residential Community Structure

Excluding industrial sectors and undeveloped peripheral lands to focus exclusively on residential communities reveals an even starker pattern. For this analysis, only residential communities not currently undergoing active development (Building Out) were included. When examining residential neighborhoods, the share of the Solid class increases to 63.24%.

The composition of residential areas is distributed as follows:

• Solid: 63.24%
• Park: 19.28%
• Lawn: 13.82%
• Forest: 1.91%
• Water: 1.74%

The distribution of cover types across residential communities demonstrates that for the overwhelming majority of neighborhoods, the median area share of impervious surfaces falls within the 60–65% range. Grassy cover (Lawn), which residents often perceive as a primary “green environmental amenity”, accounts for only 13.82% in residential areas. Under Calgary’s dry continental summer conditions, unirrigated lawn rapidly enters moisture deficit and loses its capacity to cool the air via evapotranspiration.

Concrete-to-Tree Imbalance: The Solid-to-Tree Ratio Index

To evaluate the ecological and microclimatic quality of the urban environment, the Solid-to-Tree Ratio index was calculated—representing the ratio of non-vegetated area to total tree canopy area ($\frac{\text{Solid}}{\text{Forest} + \text{Park}}$).

The spatial distribution map of this index (on a logarithmic scale) clearly illustrates Calgary’s microclimatic polarization:

• Critical Shade Deficit (Solid-to-Tree Ratio > 15–50): Downtown Commercial Core (Ratio = 56.2): for 110.8 ha of artificial surfaces, there are only 0.41 ha of forest and 1.56 ha of parkland trees. Beltline (Ratio = 24.6): high development density coupled with minimal tree canopy presence. New Suburban Communities on the Periphery (Redstone — 23.8, Seton — 20.8, Rangeview — 16.8, Martindale — 12.4): modern high-density single-family development standards featuring narrow lots and minimal yard space, combined with neighborhood youth, result in an almost complete absence of mature tree canopy.
• Ecological Buffers (Solid-to-Tree Ratio < 1.0): Communities adjacent to river valleys and natural ravines (Discovery Ridge — 0.5, Queens Park Village — 0.3, Wildwood — 0.6, Roxboro — 0.5, Elbow Park — 0.9, Upper Mount Royal — 0.8). In these communities, tree canopy area exceeds the exposed area of asphalt and roofs, generating a sustained shading and cooling effect.

All area metrics for land cover types and the Solid-to-Tree Ratio within residential communities are compiled in the interactive table below (hectares). You can sort metrics in ascending or descending order. To quickly find your community, start typing its name in the search box (Пошук – Search).

Community	Forest	Lawn	Park	Solid	Water	Solid-to-Tree Ratio
ABBEYDALE	2,15	25,18	29,84	109,23	4,03	3,4
ACADIA	2,54	40,78	62,39	284,01	0,46	4,4
ALBERT PARK/RADISSON HEIGHTS	0,67	34,61	37,72	170,92	7,01	4,5
ALTADORE	4,34	24,55	45,96	135,83	2,5	2,7
APPLEWOOD PARK	0,94	28	18,92	105,64	2,01	5,3
ARBOUR LAKE	2,73	65,66	61,76	306,04	3,27	4,7
AUBURN BAY	8,3	59,34	60,79	297,45	26,75	4,3
BANFF TRAIL	0,15	7,11	28,22	109,97	0,04	3,9
BANKVIEW	0,19	0,87	13,32	54,32	0	4
BAYVIEW	1,28	1,93	13,32	25,8	0,03	1,8
BEDDINGTON HEIGHTS	0,84	25,6	54,21	239,48	0	4,4
BEL-AIRE	1,71	0,31	13,03	18,48	0,18	1,3
BELTLINE	1,18	2,18	9,98	274,61	6,47	24,6
BONAVISTA DOWNS	0,03	14,54	7,28	35,34	0	4,8
BOWNESS	34,67	54,2	159,2	312,35	1,9	1,6
BRAESIDE	0,75	16,16	40,83	138,29	0,01	3,3
BRENTWOOD	1,66	19,39	58,81	209,21	0,17	3,5
BRIDGELAND/RIVERSIDE	6	29,35	66,97	169,84	32,96	2,3
BRIDLEWOOD	1,62	68,86	43,86	195,24	4,52	4,3
BRITANNIA	0,47	2,58	16,61	30,52	0,03	1,8
CAMBRIAN HEIGHTS	0,12	6,31	23,34	56,59	0	2,4
CANYON MEADOWS	0,97	30,26	71,53	205,27	0,02	2,8
CAPITOL HILL	4,33	5,97	32,15	98,94	0,12	2,7
CASTLERIDGE	0,18	6,64	16,12	98,32	0	6
CEDARBRAE	0,63	28,2	35,44	138,75	0,02	3,8
CHAPARRAL	8,61	177,79	73,61	263,75	21,15	3,2
CHARLESWOOD	1,98	24,79	55,15	101,19	0	1,8
CHINATOWN	1,28	0,26	3,52	18,16	1,03	3,8
CHINOOK PARK	0,09	0,69	19,05	38,39	0	2
CHRISTIE PARK	6,13	3,8	20,82	51,15	0,03	1,9
CITADEL	2,4	35,25	56,38	179,27	0	3
CLIFF BUNGALOW	1,26	1,89	7,14	25,87	0,17	3,1
COACH HILL	0,87	9,26	29,05	68,24	0,01	2,3
COLLINGWOOD	3,65	35,06	40,82	76,27	0,49	1,7
CORAL SPRINGS	2,84	46,03	21,71	98,79	9,07	4
CORNERSTONE	13,43	305,5	25,13	379,97	40,6	9,9
COUGAR RIDGE	27,93	24,36	57,66	115,05	1,3	1,3
COUNTRY HILLS	3,15	66,26	22,59	94,37	2,57	3,7
COUNTRY HILLS VILLAGE	0,11	14,67	5,06	68,41	7	13,2
COVENTRY HILLS	0,11	50,31	43,46	314,71	0	7,2
CRESCENT HEIGHTS	0,8	4,59	50,78	99,67	6,04	1,9
CRESTMONT	39,68	105,64	43,26	78,69	2,39	0,9
DALHOUSIE	4,02	32,26	85,58	212,5	0,15	2,4
DEER RIDGE	0,35	20,55	25,4	95,4	0,01	3,7
DEER RUN	1,17	27,31	45,52	117,89	0,08	2,5
DIAMOND COVE	1,78	6,73	17,68	27,2	0,15	1,4
DISCOVERY RIDGE	94,72	45,12	109,58	101,34	6,99	0,5
DOUGLASDALE/GLEN	11,53	73,94	169,34	354,77	34,64	2
DOVER	6,3	66,3	83,58	238,66	20,5	2,7
DOWNTOWN COMMERCIAL CORE	0,41	0,37	1,56	110,81	14,79	56,2
DOWNTOWN EAST VILLAGE	0,37	4,99	3,95	42,11	1,18	9,7
DOWNTOWN WEST END	0,48	0,97	2,33	31,32	1,08	11,1
EAGLE RIDGE	0,79	2,14	11,74	22,95	0,06	1,8
EAU CLAIRE	2,71	1,15	15,78	27,17	4,28	1,5
EDGEMONT	18,07	102,92	182,74	358,08	0,28	1,8
ELBOW PARK	14,95	5,75	75,02	77,66	5,56	0,9
ELBOYA	1,38	5,32	22,4	43,02	0,08	1,8
ERIN WOODS	0,34	17,01	17,73	124,13	0,78	6,9
ERLTON	0,87	4,88	18,84	25,95	0	1,3
EVANSTON	1,52	87,66	52,2	342,94	3,9	6,4
EVERGREEN	3,96	53,48	80,78	349,19	3,26	4,1
FAIRVIEW	0,03	6,13	34,8	91,13	0	2,6
FALCONRIDGE	0,03	23,03	22,27	169,07	0	7,6
FOREST HEIGHTS	0,04	11,37	31,65	107,48	0	3,4
FOREST LAWN	0,04	22,51	34,77	174,06	0,02	5
GARRISON WOODS	0,93	2,71	15,28	58,8	0,05	3,6
GLAMORGAN	0,21	20,08	42,89	140,26	0,01	3,3
GLENBROOK	0,39	15,93	41,84	137,88	0,03	3,3
GLENDALE	0,45	14,39	46,7	75,46	0	1,6
GREENVIEW	0,17	5,93	8,98	30,5	0	3,3
GREENWOOD/GREENBRIAR	6,74	12,74	16,66	82,87	0,03	3,5
HAMPTONS	9,17	71,07	84,52	178,86	6,58	1,9
HARVEST HILLS	1,09	18,68	31,85	161,83	4,45	4,9
HAWKWOOD	3,07	24,64	76,75	216,44	0,11	2,7
HAYSBORO	0,09	14,15	58,58	193,69	0,33	3,3
HIDDEN VALLEY	3,53	99,48	89,71	233,56	4,46	2,5
HIGHLAND PARK	0,63	15,16	36,99	84,06	0	2,2
HIGHWOOD	0,21	6,95	32,51	55,38	0,01	1,7
HILLHURST	2,24	12,51	44,85	136,58	10,72	2,9
HOUNSFIELD HEIGHTS/BRIAR HILL	0,62	3,8	31,16	82,39	0,02	2,6
HUNTINGTON HILLS	0,39	75,47	75,01	326,16	0	4,3
HUXLEY	2,57	149,92	48,79	208,1	8,96	4,1
INGLEWOOD	7,35	43,27	79,42	118,1	1,97	1,4
KELVIN GROVE	0,24	5,23	20,16	51,38	0,03	2,5
KILLARNEY/GLENGARRY	0,93	6,07	30,58	146,88	0,03	4,7
KINCORA	1,39	69,76	36,62	117,65	1,78	3,1
KINGSLAND	0,01	7,79	25,57	100,47	0,02	3,9
LAKE BONAVISTA	3,6	52,43	96,38	343,8	30,12	3,4
LAKEVIEW	2,65	16,17	53,26	151,95	0,05	2,7
LINCOLN PARK	0,15	22,17	17,16	108,87	0,16	6,3
LOWER MOUNT ROYAL	0,09	0	1,82	26,51	0,03	13,9
MACEWAN GLEN	0,35	8,47	30,17	102,89	0	3,4
MAHOGANY	3,17	93,4	44,05	459,57	29,96	9,7
MANCHESTER	0,04	0,36	2,47	44,4	0,04	17,7
MAPLE RIDGE	3,53	75,97	54,75	134,35	1,02	2,3
MARLBOROUGH	0,02	24,36	29,21	212,11	0	7,3
MARLBOROUGH PARK	0,05	27,73	26,3	194,82	0	7,4
MARTINDALE	0,15	24,69	18,01	226,08	0	12,4
MAYFAIR	0,22	0,4	13,95	14,39	0,01	1
MAYLAND HEIGHTS	0,14	29,18	34,31	138,69	0	4
MCKENZIE LAKE	9,48	41,59	128,42	293,03	34,33	2,1
MCKENZIE TOWNE	8,25	52,31	64,42	344,36	13,23	4,7
MEADOWLARK PARK	0,1	3,15	11,47	48,58	0	4,2
MEDICINE HILL	24,05	18,53	33,67	43,45	0,03	0,8
MIDNAPORE	1,6	28,13	62,4	192,9	10,6	3
MILLRISE	0,27	9,15	27,91	140,99	0	5
MISSION	1,46	0,33	7,51	42,54	1,68	4,7
MONTEREY PARK	2,95	84,56	47,89	204,11	8,79	4
MONTGOMERY	2,9	44,73	59,3	157,29	31,05	2,5
MOUNT PLEASANT	1,86	11,28	55,82	118,47	0,02	2,1
NEW BRIGHTON	3,64	28,75	34,01	208,98	9,5	5,6
NOLAN HILL	1,99	30,86	31,85	137,59	2,23	4,1
NORTH GLENMORE PARK	0,24	8,76	24,65	81,98	0	3,3
NORTH HAVEN	0,02	7,18	15,17	62,16	0	4,1
NORTH HAVEN UPPER	0,04	2,54	3,26	21,18	0	6,4
OAKRIDGE	1,36	20,97	71,01	147,94	0	2
OGDEN	5,48	55,77	122,37	213,21	17,24	1,7
OSPREY HILL	1,59	2,14	3,61	13,82	0	2,7
PALLISER	0,14	5,92	19,85	72,8	0,06	3,6
PANORAMA HILLS	5,58	111,33	118,25	392,76	5,85	3,2
PARKDALE	0,97	6,3	17,64	72,22	7,98	3,9
PARKHILL	2,15	3,36	20,37	39,69	0,05	1,8
PARKLAND	0,79	26,66	46,94	114,56	0	2,4
PATTERSON	14,28	26,87	64,69	98,75	0	1,3
PENBROOKE MEADOWS	0	16,95	22,32	164,65	0	7,4
PINERIDGE	0,06	29,92	36,89	197,19	0	5,3
POINT MCKAY	2,22	2,55	12,52	14,57	6,44	1
PUMP HILL	2,46	10,08	30,23	51,44	0,09	1,6
QUEENS PARK VILLAGE	0,4	26,15	28,11	7,32	0	0,3
QUEENSLAND	0,12	16,67	24,12	114,89	0	4,7
RAMSAY	0,52	7,51	23,86	66,71	0,01	2,7
RANCHLANDS	4,74	19,04	60,72	149,54	0	2,3
RANGEVIEW	1,15	148,99	17,32	310,49	4,54	16,8
RED CARPET	0	10,17	6,94	39,18	0	5,6
REDSTONE	0,48	31,01	6,08	155,99	1,86	23,8
RENFREW	1,85	49,69	65,91	148,32	0,06	2,2
RICHMOND	0,56	10,38	41,33	138,29	0,02	3,3
RIDEAU PARK	2,25	0,79	12,79	10,8	0,09	0,7
RIVERBEND	16,41	28,43	84,93	239,13	37,41	2,4
ROCKY RIDGE	8,69	30,93	68,96	167,55	4,76	2,2
ROSEDALE	0,77	2,28	27,43	35,74	0	1,3
ROSEMONT	0,14	1,38	11,37	32,07	0	2,8
ROSSCARROCK	0,03	6,52	10,99	89,21	0	8,1
ROXBORO	2,24	0,12	16,21	8,67	0,03	0,5
ROYAL OAK	10,57	56,05	81,03	201,72	6,87	2,2
RUNDLE	0,23	21,67	35,59	206,71	0	5,8
RUTLAND PARK	0,75	3,49	15,45	48,46	0	3
SADDLE RIDGE	3,71	85,32	36,14	422,11	12,39	10,6
SANDSTONE VALLEY	0,32	13,51	34,09	132,16	0	3,8
SCARBORO	0,66	2,85	18,66	23,42	0	1,2
SCARBORO/SUNALTA WEST	0,31	4,36	9,53	21,27	0,01	2,2
SCENIC ACRES	3,99	66,87	108,64	249,09	14,97	2,2
SETON	0,98	60,84	16,94	372,83	4,56	20,8
SHAGANAPPI	10,36	31,49	51,38	54,04	0,84	0,9
SHAWNEE SLOPES	0,95	8,81	22,3	92,91	0,13	4
SHAWNESSY	1,38	26,41	52,92	279,36	4,83	5,1
SHERWOOD	1,62	62,81	28,07	143,6	3,25	4,8
SIGNAL HILL	3,37	72,51	111,6	356,02	0,48	3,1
SILVER SPRINGS	10,63	67,93	134,18	264,35	24,69	1,8
SKYVIEW RANCH	0,8	34,77	13,65	178,39	3,54	12,3
SOMERSET	0,74	30,36	36,79	148,92	1,84	4
SOUTH CALGARY	0,38	2,11	11,98	73,74	0,07	6
SOUTHVIEW	1,68	27,43	43,01	76,36	10,89	1,7
SOUTHWOOD	0,08	25,62	35,83	204,97	0,12	5,7
SPRUCE CLIFF	12,51	6,35	32,32	56,01	0,3	1,2
ST, ANDREWS HEIGHTS	0,36	12,77	25,06	77,43	0,49	3
STRATHCONA PARK	14,28	27,74	74,84	153,73	0,1	1,7
SUNALTA	0,46	5,24	11,04	75,86	0,61	6,6
SUNDANCE	2,18	54,42	63,05	265,66	12,84	4,1
SUNNYSIDE	1,38	2,78	28,25	57,6	12,22	1,9
SYMONS VALLEY RANCH	0	2,19	0,75	12	0	16
TARADALE	1,5	44,99	18,29	214,62	5,22	10,8
TEMPLE	0,02	25,86	24,93	212,99	0	8,5
THORNCLIFFE	0,37	44,39	86,54	195,63	0,02	2,3
TUSCANY	23,01	124,14	166,87	362,26	3,62	1,9
TUXEDO PARK	0,16	3,53	26,65	102,55	0,01	3,8
UNIVERSITY DISTRICT	0,53	19,12	14,92	80,08	0,75	5,2
UNIVERSITY HEIGHTS	0,59	3,4	26,03	51,66	0,03	1,9
UNIVERSITY OF CALGARY	1,26	15,08	32,93	100,83	0	2,9
UPPER MOUNT ROYAL	7,04	3,7	61,3	55,1	0,31	0,8
VALLEY RIDGE	38,68	60,24	90,41	134,47	2,94	1
VARSITY	22,09	80,33	228,36	337,03	13,43	1,3
VISTA HEIGHTS	0,01	33,3	17,14	63,5	0	3,7
WALDEN	5,32	47,33	25,34	165,03	3,01	5,4
WEST HILLHURST	2,29	14,76	43,81	155,52	19,61	3,4
WESTGATE	0,07	7,39	29,18	79,78	0	2,7
WHITEHORN	0,08	23,62	40,64	199,94	0	4,9
WILDWOOD	46,24	31,28	95,64	82,93	0,05	0,6
WILLOW PARK	9,28	39,57	80,21	210,24	0,79	2,3
WINDSOR PARK	16,63	17,32	32,54	60,06	0,54	1,2
WINSTON HEIGHTS/MOUNTVIEW	6,8	110,19	63,18	121,02	0,88	1,7
WOODBINE	1,03	41,1	57,22	223,36	0,03	3,8
WOODLANDS	5,24	66,45	68,51	138,72	1,63	1,9

Conclusions

Calgary is visually perceived as a green city due to the spatial prevalence of lawns and open turfgrass (Lawn — 28.6%). However, from a thermodynamic perspective and in terms of mitigating urban heat island effects, lawns cannot substitute for tree canopy cover.

In new residential areas, non-vegetated surfaces account for 60–75% of total area, while the ratio of artificial surfaces to tree cover reaches extremely high levels (15–25). This establishes long-term microclimatic risks for residents of these communities. The 10 m resolution digital LULC map created in this study enables precise localization of “shade deficit hotspots” down to the level of individual streets and residential lots, supporting targeted urban forestry programs (Urban Canopy Expansion).

Technical Appendix

Classification utilized the ground data and synthesis of optical and synthetic aperture radar (SAR) satellite data acquired during the summer period (June–August) 2025 over Calgary and its immediate surroundings:

• Sentinel-1 SAR (European Space Agency): over 430 million individual measurements. Synthetic Aperture Radar data in VV/VH polarizations and derived indices at 10 m spatial resolution. The radar signal is critical for separating smooth artificial surfaces (asphalt, rooftops) from the rough structure of tree canopies.
• Sentinel-2 (European Space Agency): over 4.1 billion individual spectral measurements (4,192 × 3,423 pixels across a 287-band time-series and derived spectral feature stack at 10 m spatial resolution). Multispectral bands and derived indices at 10 m spatial resolution.
• Ground Truth: Field training and validation samples were collected during extensive walking surveys across various city quadrants, followed by verification using ultra-high-resolution aerial photography. The total area of training and validation polygons for the current machine learning model version (v6.0) is 685 hectares.

The machine learning workflow was built in R using the terra, tidyverse, ranger, and caret packages. Visualizations were generated using ggplot2. Spatial boundaries and community attributes were obtained from the Community Boundaries dataset (City of Calgary Open Data).

Model performance evaluation on an independent test dataset demonstrated high reliability:

• Overall Accuracy: 98.29% (95% CI: 98.16% – 98.41%). Cohen’s Kappa: 0.9691
• Class-wise Accuracy Statistics (Positive Predictive Value / Precision): Solid: 99.98% (The model yields virtually no false positives for concrete and asphalt); Water: 99.95%; Lawn: 99.50%; Forest: 97.59%; Park: 56.56%

The lower Positive Predictive Value (PPV) for the Park class (56.56%) is a natural outcome of the mixed pixel effect at a 10×10 m spatial resolution. Standalone trees or sparse tree clusters have canopy dimensions comparable to or smaller than a single Sentinel pixel (100 m²). Consequently, the spectral and radar response of the pixel represents a mixed signal from foliage, understory turfgrass, and cast shadow, leading to partial misclassification with the Lawn and Forest classes.

Methodological Limitations. The developed model operates at a spatial resolution of 10 × 10 m. This scale is designed for meso- and macro-level analysis of the urban climate structure and incorporates the following characteristics:

• Dominant Cover: Each pixel is classified according to its dominant optical and radar response. Small land cover features substantially smaller than 100 m² (young saplings, isolated trees with sparse canopies, as well as narrow footpaths, yard infrastructure, and small structures) do not form an independent spectral signature and are integrated into the dominant class of the cell.
• Condition and Health Sensitivity: Trees with damaged or senescent canopies (reduced Leaf Area Index, LAI) exhibit a diminished response in the near-infrared and shortwave infrared spectrum (NIR/SWIR). From a microclimatic perspective (evapotranspirational cooling and shading), their thermal behavior resembles low vegetation (Lawn), which accurately reflects their actual contribution to the thermal balance of the area.
• Scope of Application: This map represents a model of dominant land cover types for urban heat island assessment and does not substitute for a high-resolution Arboricultural Tree Inventory, which requires aerial survey data at <0.5 m resolution or terrestrial/airborne LiDAR scanning.

Disclaimer: This independent study is conducted for scientific and educational purposes (CC BY 4.0). The maps and conclusions are based on remote sensing and machine learning methods and do not constitute an official engineering opinion. Consult with a practicing Professional Member of APEGA before making design, construction, or public health decisions.