Urban forests provide an array of social, environmental, and economic benefits to people living in the urban areas. According to the Forest Service of the U.S. Department of Agriculture (USDA), urban forests are made up of individual trees, urban parks, river corridors, wetlands, nature preserves, gardens, landscaped boulevards, greenways, river, and coastal promenades [1].

These green spaces have the potential of providing ecosystem benefits to the human population as well. These ecosystem benefits include microclimate regulation, rainwater drainage, sewage treatment, air filtration, noise reduction, recreational and cultural values [2] [3]. In essence, these ecosystem services improve the quality of the air, increase carbon sequestration, and reduce the urban heat island effect.

Increasingly, in the past decade, urban forest ecosystem has come under intense pressure because of the threat posed by multiple factors. Flooding caused by natural disaster and anthropogenic activities have exerted a lot of pressure on urban forest ecosystem. Increased rainfall associated with climate change has altered the flooding experienced in many urban areas of the world including the United States. Recent widespread flooding across Louisiana during the Hurricanes Katrina and Idai has highlighted the significant negative impacts that flooding can have on Urban forest ecosystem. Baton Rouge, Louisiana is not an exception [4] [5].

With all the benefits of urban forests, it is highly essential to monitor the extent and health of these green spaces over time and report on their status and trends as well. This helps to generate information to support decision making in terms of the condition of these trees as well as their extent in Louisiana or the US at large. Irrespective of how small these urban forests such as parks, street trees and landscaped boulevards are, they all have a role to play in the ecosystem such as the regulation of microclimate, filtration of the air and the management of stormwater. Details of these individual trees and green spaces might not be entirely visible in medium and low-resolution satellite imagery and hence might not fully represent what is on ground. This can affect biomass estimates, carbon sequestration and other land cover analysis. Remote sensing has played an important role in monitoring condition of the urban forest ecosystem [6] [7] [8].

Knowing the condition and health of urban forest structure are some of the keys to their sustainability. Information on the physical characteristics of the urban trees such as tree species composition, species diversity, leaf area, biomass, tree density, number of trees and tree health are useful parameters in urban forest monitoring [9]. Thus, this study uses fine resolution image fusion techniques to improve the identification and monitor urban forest in Baton Rouge.

Earth observation technology plays a major role in identifying, measuring and monitoring urban forests as well as detailed data on the location. This technology leverages on the power of remote sensing and space systems to capture and store information on the various thematic areas at different energy levels known as bands. Remote sensing also offers a wide range of temporal, spectral, radiometric, and spatial resolution to support analysis of features on the earth surface [10]. Early work of [11] reveals that, the integration of a high-resolution multispectral satellite imagery for urban forest monitoring ensures that more details of such trees are captured to increase the accuracy levels during the image analysis process.

Fonseca et al. [10] further explains that image fusion is useful in areas that have complex morphological structure including urban areas, heterogeneous forested areas, landslide scars, and classifying intra-urban land cover. Indeed, several research papers on multi-sensor image fusion have studied 1) the mapping of urban tree cover by fusing multispectral image and LiDAR point cloud data to improve point cloud classification quality, and consequently to produce tree canopy cover maps [12], 2) the fusion of different dataset such as satellite imagery from the same sensor in Turkey’s third largest city of Izmir using Principal Component, IHS, and Brovey Transform techniques to improve the spatial resolution of Landsat 7 ETM imagery used in the study. Results indicate, improved spatial resolution of fused Landsat 7 ETM imagery using Principal Component and Brovey Transform techniques [13], 3) the use of Satellite imagery fusion methods such as intensity-hue-saturation (IHS), principal component analysis (PCA) and high pass filter (HPF) to improve urban land cover classes identification [11], and 4) the applications of high-resolution panchromatic and multispectral Quickbird images, and TerraSAR synthetic aperture radar (SAR) data are used for urban land use change studies in the central part of Ulaanbaatar, the capital city of Mongolia. Results of the fused image in Ulaanbaatar city show an increased urbanization leading to land use change [14]. [15] fused airborne LiDAR data with aerial images for a tree-level urban forest inventory. The aboveground points, vegetation points were isolated from non-vegetation points. This was achieved by using the Normalized Difference Vegetation Index (NDVI) derived from the hyperspectral data. Results illustrated that individual trees can be detected from LiDAR point cloud data using the proposed tree climbing algorithm, and tree crowns can be delineated using the donut expanding and sliding method. [16] fused LiDAR and Landsat Thematic Mapper data to discriminate land covers in the urban areas of Charlotte, North Carolina, USA by using supervised maximum likelihood (ML) and classification tree (CT) methods. The authors found that fusion of higher-resolution structural and intensity surface models from LiDAR with TM data improved total and class-level classification accuracies. Also, in the suburb of Melbourne, Australia, Parmehr et al. [17] fused airborne LiDAR and optical remote sensing data to map urban tree cover. Results showed an improved reliability of LiDAR point cloud classification of tree cover.

In all of these studies, improving low spatial resolution multispectral imagery with very fine resolution panchromatic images as well as enhancing and improving visualization of spatial features has been a major goal. In Louisiana, studies focusing on fusing high resolution imagery with low spatial resolution multispectral imagery for urban forestry studies at specific city are very limited. In 2003, Louisiana Department of Environmental Quality (LDEQ) conducted Louisiana-wide study and utilized 2002 Landsat Thematic Mapper Plus (30 m resolution) RGB 7-5-3 fused with higher resolution (15 m) panchromatic band in order to improve visualization of spatial features of the entire Parish [18]. It is important to know that, LDEQ study did not focus on specific city in the state. Also, fusing of the images used only Landsat ETM+ band 8 data with 15 m resolution. To this end, the main objective of this study adopts remote sensing techniques, specifically, using very fine resolution image fusion to improve the identification and monitor urban forest in Baton Rouge. Fusing of high-resolution multispectral images at different resolutions will be performed to assist policymakers in the effective planning and management of urban forest ecosystem in Baton Rouge.

Figure 3 and Figure 4 show band combination of Landsat 8 satellite image and PlanetScope image. Figure 5: NIR band of the PlanetScope image.

Figure 6 shows the GPS location and distribution of urban forestry parks in East Baton Rouge Parish. The road network indicates that these parks are situated close to the various cities and are easily accessible to the community members.

Figure 7 shows a comparison of the fused Landsat 8 and PlanetScope image on the left and the 30 m Landsat 8 image on the right.

The image captured highlights the Northwestern part of East Baton Rouge. The fused image reveals more details of the landscape compared to the image on the right. Moreover, features in the right image are more pixelated compared to the left image. The parks, trees and greenspaces are difficult to identify in the Landsat 8 image on the right. This is same for Figures 8-11.

Overall, the fused images using Intensity-Hue-Saturation (IHS) method have better contrast and improve visualization of spatial features than non-fused images. An understanding and the use of image fusion can facilitate better decisions and policymaking. Image fusion generates a high-resolution output that highlights lots of details within the landscape. This technique can identify all manner of changes and is more effective than existing land-use land cover change detection [30]. Image fusion provides city planners and land-use experts with relevant information on the location of facilities and resources in the cities. This technique helps decision-makers understand and appreciate the complexities of urban development through the lens of remote sensing and GIS and guides them to propose strategies that ensure that urban areas remain resilient in a changing climate. In other words, researchers and various stakeholders can extract more reliable detail to assess land use land cover changes for better governmental

decision-making eliminating redundancy.

Louisiana is known for various natural disasters including hurricanes and flooding. The precision of these fused images can accurately support emergency response and urban forest management such as tree debris removal after a major hurricane incidence. While such information can be used as decision support tool for urban forest managers and those charged with policy making in the years ahead, it can also serve as educational tool to guide stakeholders in the sustainable management of urban trees. Policy recommendation in form sequential measurement of urban forest over time can help to track changes and allows for better informed policy and decision making with respect to urban forest management.

The authors would like to acknowledge the United States Department of Agriculture (USDA) National Institute of Food and Agriculture (NIFA) McIntire Stennis Forestry Research Program funded project with award number NI22MSCFRXXXG077. Also, we would like to extend our sincere gratitude to the Dean of Graduate Studies at Southern University in Baton Rouge, Louisiana, Professor Ashagre Yigletu for providing graduate assistantships to some of the graduate students on this paper in order to promote research and help the students acquire the necessary skill development while earning a graduate degree.

The authors declare no conflicts of interest regarding the publication of this paper.

Twumasi, Y.A., Asare-Ansah, A.B., Merem, E.C., Loh, P.M., Namwamba, J.B., Ning, Z.H., Yeboah, H.B., Anokye, M., Armah, R.N.D., Apraku, C.Y., Atayi, J., Frimpong, D.B., Okwemba, R., Oppong, J., Kangwana, L.A., Mjema, J., Njeri, L.W., McClendon-Peralta, J. and Jeruto, V. (2022) Fusion of Landsat 8 OLI and PlanetScope Images for Urban Forest Management in Baton Rouge, Louisiana. Journal of Geographic Information System, 14, 444-461. https://doi.org/10.4236/jgis.2022.145024