Abstract
The main purpose of this research is to model the spatial distribution of carbon sequestration, CO2 absorption, and oxygen production by trees within Isfahan city, Iran, in 2020. To quantify carbon sequestration, we accessed a sample group of trees with measured biophysical attributes. First, we calculated the biomass and carbon sequestration of a tree using the allometric and photosynthesis equations. Then, to model the spatial distribution of carbon sequestration, we used Geographic Weighted Regression (GWR) method. In this model, the amount of calculated carbon sequestration was the dependent variable, whereas the difference between vegetation index of ΔExGR(Excess Green Plant Index minus Excess Red Plant Index)from the Worldview image was the independent variable. Subsequently, the spatial distribution map of CO2absorption and oxygen production was generated. The total value of annual carbon sequestration, CO2 absorption, and O2 production was about 7704.22, 28274.502, and 20570.16 tons, respectively. The results showed that there was a strong correlation between the ΔExGR index of the canopy with calculated carbon. Integrating the ΔExGR index from a high-resolution image with calculated carbon can contribute to developing a fast, accurate, and low-cost method in estimating carbon sequestration and modeling its spatial distribution in urban areas. In conclusion, the results of this research can be implemented by land use planners in order to integrate urban ecosystem service concept (i.e., carbon sequestration) in planning process towards sustainability of the cities.
Key Word: Urban Ecosystem Service; Urban Ecology; Climate Regulation Service; Green Infrastructure; Geographic Weighted Regression (GWR);
Introduction
Increasing fossil fuel consumption in urban areas due to population growth brings about a large amount of greenhouse gases emission into the atmosphere (Houghton, 2001). Among greenhouse gases, carbon dioxide plays a significant role in global warming (Lal, 2004; Peters, 2001; Petit et al., 1999; Scott et al., 2002). In climate change studies, urban areas attributes to emit a high proportion of carbon dioxide into the atmosphere (Churkina, 2008, Grimm et al., 2008).
Urban green areas (lawns and trees) provide multitude of ecosystem services like climate regulation and air purification. Carbon sequestration service by vegetation cover is a well-recognized urban ecosystem service mitigating the atmosphere’s carbon dioxide (Baró et al., 2014, Gómez-Baggethun et al., 2013, Haase et al., 2014, Kiss et al., 2015, Larondelle and Haase, 2013).
Carbon sequestration refers to a process in which the atmospheric carbon dioxide is converted into organic compounds by photosynthesis process in trees, plants, phytoplankton, and algae (Adams et al., 1990, Nanda et al., 2016, NAYAK et al., 2020, Tornquist et al., 2009). Carbon sequestration amount is related to the growth rate, species type, and age of the tree (McPherson 1998). During the photosynthesis process, CO2 stores in the form of cellulose. Also, the other portion of the carbon transfers to the soil in organic form (Dwivedi et al., 2009, Komiyama et al., 2005, MacFarlane, 2009, Miller et al., 2015, Nowak and Crane, 2002, Nowak and Dwyer, 2007, Rowntree and Nowak, 1991, Tang and Li, 2013, Ward Thompson et al., 2016, Zirkle et al., 2012). The resources of carbon storage in an ecosystem include above-ground and below-ground biomass, litter and plant residues, and soil organic matter (Nowak and Crane, 2002, Sinoga et al., 2012).
Additionally, green spaces are considered as oxygen production resources in urban areas. Oxygen production of plants directly related to the carbon storage process. Estimating produced oxygen and carbon sequestration by vegetation in an urban area is essential in dealing with air pollution (Nowak et al., 2007).
Considering previous literature related to tree biomass estimation and carbon sequestration, they can be divided into two general categories: the first body of research is characterized by ground sampling or measuring biological variables in a laboratory environment. Numerous studies have been done in this area, which included: estimating carbon storage in biomass in a forested area in Chile (Espinosa et al., 2005), biomass estimation and leaf area index in mangrove forests of Japan ( Khan et al., 2005), estimating the biomass of ten tree species in temperate forests of China (Wang, 2006), calculating soil biomass of mangrove species in Brazil (Medeiros and Sampaio, 2008), and estimating above-ground biomass and carbon sequestration in rainforests in Thailand (Terakunpisut et al., 2007). Other studies in this field included the research by Aguaron and McPherson, 2012, Bernal et al., 2018, Nowak et al., 2013, Tor-ngern and Leksungnoen, 2020, Townsend ‐ Small and Czimczik, 2010, Velasco et al., 2016. The second class of research is organized based on satellite image and remote sensing techniques to measure the biomass of the plants. For instance, estimating the amount of biomass of Acacia species, silver cypress, berry tree using linear regression model and applying Quick bird data and NDVI and DVI indices in Isfahan, Iran (Hosseini et al., 2015). Mirrajabi and colleagues in 2016 estimated biomass of broadleaf and coniferous species using GeoEye images in Chitgar park, Tehran, Iran (Mirrajabi et al., 2016). Amini and Sadeghi benefitted from ALOS data and multiple regression equations to estimate the amount of forest biomass (Aliabadi and Entezari, 2014). The other research in this category can be found in related papers done by Deng et al., 2011, Günlü et al., 2014, Hall et al., 2006, Raciti et al., 2014, Strunk et al., 2014.
Previous studies show that allometric equations were used to determine the above-ground and below-ground biomass of trees as well as to estimate the amount of carbon storage.
These equations consider several parameters like diameter at breast height (DBH), tree height, and wood density to estimate biomass in a single tree unit (Aguaron and McPherson, 2012).
In this research, we used allometric equations to calculate the biomass of a tree. Then to model the spatial distribution of carbon sequestration, we integrated the results of the allometric equation with the spectral data of the satellite image.
Another influential factor in the accuracy of estimating carbon storage is statistical modeling methods. A wide range of methods, including parametric, semi-parametric, and non-parametric methods, have been used to quantify carbon storage using remote sensing (Wu et al., 2016)
In a cumulative body of research, different statistic methods were used to integrate ground data (from the allometric equation) into remotely-sensed data to quantify various characteristic of trees like a canopy, biomass, and carbon storage (Carreiras et al., 2006, Cartus et al., 2012, Cutler et al., 2012, Ghanbari Motlagh et al., 2020, Hamdan et al., 2015, Lucas et al., 2010, Moradi et al., 2018, Mutanga et al., 2012, Raciti et al., 2014, Shataee et al., 2012).
To the best of our knowledge, to estimate carbon sequestration, the previous studies measured several trees as samples in a plot to examine the relationship between ground data and satellite data through the linear regression (Deng et al., 2011, Mirrajabi et al., 2016, Raciti et al., 2014). However, in this study, for the first time, we used geographic weighted regression (GWR) to create a relationship between spectral data of each tree (canopy reflection (drip line)) with the calculated ground biomass. It was assumed that integrating the remotely sensed data with the measured biomass in the GWR model can provide reliable results, contributing to addressing the spatial distribution of carbon sequestration as well as estimating the amount of carbon sequestration.
Considering the background previously discussed, therefore, this article aims to develop a model to analyze the spatial distribution of carbon sequestration, CO2 absorption, and O2production within Isfahan city. To achieve this goal, we pursued the following sub-objectives:
- Calculating the above-ground and below-ground biomass of sampled trees using the allometric equations. - Estimating carbon sequestration of sampled trees using photosynthesis equation.  - Processing the satellite image with different spectral variables to recognize the most appropriate index. - Creating a GWR model to integrate the spectral data of each tree with carbon sequestration of the tree. - Providing the spatial distribution map of carbon sequestration, CO2 absorption, and oxygen production.