Fig 1.
Urban lands in the San Francisco Bay are represented by the red polygon.
Table 1.
Spatial and temporal resolution of NAIP images consulted in the study area. DOY: Day of Year.
Table 2.
Explanatory variables analyzed to model the effect on the rate of change on large tree presence and large trees per capita and canopy cover in large trees between 2005 and 2022. The socioeconomic variables were acquired in the Urban Displacement Project (Chapple et al., 2021).
Table 3.
Confusion matrix of the model test process in San Francisco plots. TP: true positive; FP: false positives; FN: False negatives; TN: True negative.
Fig 2.
Left panel: Predicted number of large streets trees vs ground truth number of large street trees by plot.
Right panel: Predicted number of large street trees vs ground truth number of large street trees by plot vs the correction with the statistical model. Dash line indicates variability.
Fig 3.
Left: Variation of the rate of change of large trees per neighborhood by year.
Right: Estimated effect of variables (large street tree density, percentage of cover, canopy area and, number of trees) on the rate of change of large trees per neighborhood by year. The dashed line indicates variability.
Fig 4.
A. Map of the normalized number of large street trees by km2 in 2005 B: Normalized total rate of change (trees per year) on large street trees by neighborhood.
C: Normalized total rate of change (trees per year) on large street trees by city. D: Normalized total rate of change (trees per year) on large street trees by county in the San Francisco Bay Area.
Fig 5.
Estimated effects of social variables on average canopy coverage per neighborhood (left panel) and the average large street trees per capita (right panel).
The dash area indicates the data variability.