www.nature.

com/scientificdata

OPEN Fine-grained urban blue-green-gray

Data Descriptor landscape dataset for 36 Chinese
cities based on deep learning
network
1 2✉
Zhiyu Xu & Shuqing Zhao

Detailed and accurate urban landscape mapping, especially for urban blue-green-gray (UBGG)
continuum, is the fundamental first step to understanding human–nature coupled urban systems.
Nevertheless, the intricate spatial heterogeneity of urban landscapes within cities and across urban
agglomerations presents challenges for large-scale and fine-grained mapping. In this study, we
generated a 3 m high-resolution UBGG landscape dataset (UBGG-3m) for 36 Chinese metropolises using
a transferable multi-scale high-resolution convolutional neural network and 336 Planet images. To
train the network for generalization, we also created a large-volume UBGG landscape sample dataset
(UBGGset) covering 2,272 km2 of urban landscape samples at 3 m resolution. The classification results
for five cities across diverse geographic regions substantiate the superior accuracy of UBGG-3m in
both visual interpretation and quantitative evaluation (with an overall accuracy of 91.2% and FWIoU
of 83.9%). Comparative analyses with existing datasets underscore the UBGG-3m’s great capability to
depict urban landscape heterogeneity, providing a wealth of new data and valuable insights into the
complex and dynamic urban environments in Chinese metropolises.

Background & Summary

Urban Landscapes are complex and dynamic geographic phenomena that are shaped by natural and human
forces1,2. These landscapes are comprised of various components, including urban green space (UGS), urban
blue space (UBS), and urban impervious surfaces (UIS), which together form the basic units of complex urban
landscape configurations3,4. UGS generally refers to vegetated land in urban areas5, such as parks, gardens, trees,
and grasses, which play a crucial role in upholding urban ecosystem equilibrium. UBS pertains to water features
within urban locales, encompassing rivers, lakes, wetlands, reservoirs, ponds, and artificial water structures3.
UIS, commonly referred to as “urban gray space”, are impervious surface features in cities caused by man-made
land use activities, such as building roofs, asphalt or concrete roads6. UGS and UBS provide multiple ecological
and social benefits, including climate regulation, water and air purification, biodiversity conservation, carbon
sequestration, recreational opportunities, and aesthetic enhancement3,7,8. However, UIS may engender negative
impacts, such as the urban heat island (UHI) effect, air quality degradation, and stormwater runoff increases6,9.
Therefore, a balanced approach to urban landscape construction necessitates the integration of all three space
types, with a focus on increasing the quantity and quality of UGS and UBS, while also minimizing the nega-
tive impacts of UIS, which can lead to a more sustainable and livable urban environment10. In the context of
global urbanization, there is a pressing need for a more precise perception of the urban landscape structure11.
Nonetheless, urban blue-green-gray (UBGG) landscapes are strikingly heterogeneous in terms of space, struc-
ture and function, as they are the outcome of dynamic interactions between biophysical and socio-economic
processes occurring at multiple spatial scales2,12. Yet, detailed and accurate UBGG landscape mapping is the
fundamental first step to understanding human–nature coupled urban systems. Therefore, it is necessary to
open the “closed box” of urban landscape structure and quantify the subtle heterogeneity of the built and natural
components within the metropolis1,13.

1
College of Urban and Environmental Sciences, Peking University, Beijing, 100871, China. 2College of Ecology and
the Environment, Hainan University, Haikou, 570228, China. ✉e-mail: [email protected]

Scientific Data | (2024) 11:266 | https://doi.org/10.1038/s41597-023-02844-2 1

www.nature.com/scientificdata/ www.nature.com/scientificdata

With the increased demand for higher resolution urban products, urban mapping products have made tre-
mendous progress toward finer scales in the last decades11,14,15. It can be attributed to the availability and acces-
sibility of very high resolution (VHR) satellite data and the support of computing platforms with substantial
computing power, such as Google Earth Engine (GEE)11. Notably, VHR imagery, with resolutions as fine as
1–3 m/pixel, has emerged as an invaluable asset for revealing urban landscapes at an increasingly detailed level
of granularity, offering a comprehensive view of the ground. Moreover, Deep Learning (DL) techniques have
emerged as a powerful tool for VHR urban landscape mapping, revolutionizing the field of intelligent classifica-
tion research in the 21st century16–18. Establishing fine-scale urban datasets for landscape/landcover interpreta-
tion and deep learning-based research has become a hot research topic in recent years15,17,19. However, present
urban landscape datasets typically focused on individual landscapes (e.g., UGS or UBS)20,21 or limited spatial
extents (usually covering several cities/provinces)22. Some scholars focused on UGS extraction to accurately
digitally twin UGS at a fine scale5,21,23, such as Brandt et al.23 utilized VHR satellite images covering more than
1.3 million km2 in West African Sahara and Sahel, detecting more than 1.8 billion individual trees in areas pre-
viously regarded as barely covered by trees. Similarly, Shi et al.5 generated 1-meter UGS maps for 31 major cities
in China using Google Earth images. Some scholars focused on UBS extraction to address the obstacles posed
by the confusion of water with heavy shadows in VHR images20,24,25, like Chen et al.25 proposed an open water
detection method in urban areas using VHR imagery, successfully identifying various types of water bodies.
Likewise, Li et al.20 proposed the water index-driven deep fully convolutional network (WIDFCN), showcasing
robustness to different shadows types and achieving high-performance water extraction in 12 test sites world-
wide. In addition, in the field of UIS extraction, scholars have also achieved notable outcomes in the applica-
tion of DL to urban building and road extraction from VHR images26–28. For example, Guo et al.28 devised a
coarse-to-fine boundary refinement network for building footprints extraction from VHR images. Nevertheless,
a limitation persists in the mapping of single landscapes or confining analyses to limited geographical extents,
failing to offer a comprehensive understanding of the highly heterogeneous interactions between human and
natural elements1,22.
Establishing an effective automatic DL model for fine-grained and large-scale UBGG dataset is a challenging
frontier in high-resolution urban landscape mapping. However, the pursuit of such datasets comes with its own
set of challenges, stemming from VHR image acquisition, manual annotation, and the intrinsic heterogeneity
of urban landscapes. First, the paramount significance of VHR imagery in capturing intricate urban landscape
details is countered by its inherent costliness and the complexities associated with its acquisition15,29. Although
Google Images has been used for some large-scale research, its restricted geographic and temporal coverage,
limited visible spectrum bands, as well as varying image quality are also inevitable drawbacks14. Second, train-
ing a UBGG network with large-scale applications and high generalization capability relies on a large-volume
sample dataset21, which poses a major challenge for nationwide landscape mapping due to the enormous data-
set, laborious annotation, and cumbersome process involved. Although some studies have proposed inno-
vative techniques employing biophysical indices or existing coarse-resolution products in conjunction with
self-supervised mechanisms to generate training labels automatically20, the label noise of resolution mismatch
of spatial resolution and the true accuracy of labels require further scrutiny. Reliable training labels are crucial
to achieving accurate fine-scale landscape mapping results but still insufficient. Third, the striking heterogeneity
characterizing the UBGG landscape at both intra- and inter-city levels and across various spatial scales presents
significant impediments to effectively mining multi-scale features1,2. The variability of urban landscapes across
geographic locations and climatic zones, such as plant type, water quality, building structure, and color, poses
significant challenges30. Additionally, mining multi-scale features from UBGG landscapes presents substantial
obstacles. Fine-scale features, encompassing spectral colors, geometrical sizes, and textural shapes, primarily
manifest in the network’s shallow layers but are often confused and invalidated at deeper levels26. Conversely,
coarse-scale features, such as global spatial context, are obtained from the deep layer but struggle to be effectively
expressed31.
China has undergone rapid development and urbanization in recent decades 32, becoming the world’s
second-largest economy. In light of this remarkable growth, a comprehensive mapping survey of large-scale and
fine-grained landscapes assumes immense significance, fostering an in-depth comprehension of urban environ-
ment, facilitating effective urban landscape management, and illuminating future development trajectories33.
Consequently, this study endeavors to develop a transferable multi-scale high-resolution convolutional neu-
ral network to generate a 3-meter resolution UBGG landscape dataset, utilizing Planet images in 36 Chinese
metropolises. Rigorous validation processes, including visual interpretation and quantitative evaluations, were
employed to assess the credibility and efficacy of the UBGG-3m dataset, further augmented by comparisons
with existing products. This dataset will enhance our understanding of fine-scale landscape distribution patterns
in Chinese metropolises, provide a deeper understanding of integrated human-nature systems from an ecologi-
cal perspective, and contribute to better urban landscape management as well as sustainable urban development
planning1,12,29.

Methods
Data collection and pre-processing. To supplement the lack of large-scale, fine-grained landscape data-
sets, this study used Planet multispectral satellite images and ancillary data to create UBGG-3m dataset. The dataset
encompasses 36 Chinese metropolises, including urban areas of 22 provincial capitals, 5 autonomous region capitals, 4
municipalities directly under the central government, and 5 municipalities with independent planning status (Fig. 1).
To account for China’s vast territorial expanse and the heterogeneity of its landforms, the 36 metropolises were divided
into four major geographic regions34: the northern region (Harbin, Changchun, Shenyang, Dalian, Beijing, Tianjin,

Scientific Data | (2024) 11:266 | https://doi.org/10.1038/s41597-023-02844-2 2

www.nature.com/scientificdata/ www.nature.com/scientificdata

Fig. 1 Spatial distribution of the 36 study cities within the four geographic regions in China.

Shijiazhuang, Taiyuan, Lanzhou, Qingdao, Jinan, Zhengzhou, and Xi’an), the southern region (Shanghai, Nanjing,
Hangzhou, Hefei, Ningbo, Wuhan, Changsha, Nanchang, Chengdu, Chongqing, Guiyang, Kunming, Nanning,
Fuzhou, Xiamen, Guangzhou, Shenzhen, and Haikou), the northwest region (Hohhot, Yinchuan, and Urumqi), and
the Qinghai-Tibet region (Xining and Lhasa).
Planet multispectral satellite images with a spatial resolution of 3 meters provide an important data source
for capturing the detailed characteristics of urban landscapes (https://www.planet.com/explorer/). Planet, with
the largest commercial Earth observation satellite constellation ever built, operates over 200 small satellites
in near-Earth orbit35. These satellites provide meter and sub-meter spatial resolution images, allowing for an
unprecedented global repeat observation frequency of once a week. This frequency and resolution enable the
capture and analysis of UBGG landscapes in unprecedented detail, providing insights into the morphology
and dynamics of the urban landscape at an unprecedented scale. In addition, the Planet multispectral imagery
comprises four bands, with the near-infrared band being particularly adept at capturing vegetation growth
information, thereby augmenting the accuracy of UGS type classification. A total of 336 clear and non-cloudy
images in summer of 2020 (June to October) were downloaded (Table 1). In cases where cloud cover obscured
images of the study area in 2020, cloud-free images from the summer of 2021 served as suitable replacements.
The Planet satellite images were preprocessed utilizing geometrical correction, image mosaic, color stretching,
band combination, and projection transformation. Finally, we obtained standard false color images covering 36
metropolitan urban areas in China, where trees and grass are shown as dark and bright red, which can be better
distinguished from UBS and UIS.
The boundaries of 36 metropolises were defined according to the administrative boundaries, which were
obtained from the Resource and Environment Science and Data Center (https://www.resdc.cn). Nonetheless,
administrative boundaries cannot distinguish between urban and rural areas, leading to potential misclassifi-
cation of urban grassland and farmland, due to their similar physical features but distinct economic attributes.
To address this challenge and improve classification accuracy, we integrated the 2018 China Urban Boundary
(CUB) data4 into our classification process. The CUB data was meticulously extracted through a human–com-
puter digitalization process from China’s Land Use/cover Dataset (CLUD), derived from Landsat images.
Notably, the CUB data is known for its high accuracy in urban boundary detection, with an overall accuracy rate
exceeding 92.65% from 2000 to 20184. Specifically, we focused on reclassifying areas outside the urban bound-
aries, ensuring that urban grasslands located outside these boundaries were accurately reclassified as farmland.
To comprehensively assess the reliability and precision of the UBGG-3m dataset, we collected several estab-
lished and widely utilized land cover datasets, as well as two high-resolution urban green space datasets for

Scientific Data | (2024) 11:266 | https://doi.org/10.1038/s41597-023-02844-2 3

www.nature.com/scientificdata/ www.nature.com/scientificdata

Province City Acquisition images date Number of images

Anhui Hefei 20210910 12
Fujian Fuzhou 20200722 2
Gansu Lanzhou 20200728, 20200802 9
Guangdong Guangzhou 20201011, 20201012 5
Guizhou Guiyang 20210606 3
Hainan Haikou 20200729, 20200824 4
Hebei Shijiazhuang 20200916 5
Heilongjiang Harbin 20200721, 20200821 8
Henan Zhengzhou 20200829 9
Hubei Wuhan 20210722, 20210724 10
Hunan Changsha 20200815, 20200816 10
Provincial capital (22)
Jiangsu Nanjing 20210730 13
Jiangxi Nanchang 20210722 7
Jilin Changchun 20200721, 20200726 15
Liaoning Shenyang 20210828, 20210829 15
Qinghai Xining 20200719, 20200726 5
Shaanxi Xi’an 20200702, 20200707 8
Shandong Jinan 20200903, 20200930 8
Shanxi Taiyuan 20210730 9
Sichuan Chengdu 20200727, 20200728 8
Yunnan Kunming 20210912 16
Zhejiang Hangzhou 20200816 3
Guangxi Nanning 20210606 9
Inner Mongolia Hohhot 20200914, 20200919 8
Ningxia Yinchuan 20210704 10
Autonomous region (5)
Tibet Lhasa 20210724 5
20200802, 20200804,
Xinjiang Uygur Urumqi 10
20200805
Beijing 20200809, 20200811 11
Chongqing 20200826 18
Municipality (4) 20210623, 20210714,
Shanghai 27
20210715
Tianjin 20200828, 20200903 9
Fujian Xiamen 20210722 1
Guangdong Shenzhen 20201026, 20201027 14
Municipalities with independent
Liaoning Dalian 20200708, 20200725 7
planning status (5)
Shandong Qingdao 20210619 8
Zhejiang Ningbo 20210714, 20210715 15

Table 1. Information of the Planet satellite images used in this study.

comparison and validation. Specifically, the land cover datasets include the 30 m GlobeLand30 in 202036, the
10 m Esri land cover in 202037, the 10 m ESA World Cover in 202038, the 1 m national-scale land-cover map
(SinoLC-1m)14. To ensure consistency with our classification system, the four land cover products were reclassi-
fied into UGS (trees, grassland and farmland), UBS, and UIS. The other two high-resolution urban green space
datasets include the 2 m Urban Tree Cover dataset (UTC-2m)21, and 1 m Urban Green Space (UGS-1m)5.

Technical framework. The workflow for generating the UBGG-3m dataset mainly includes three phases, as
depicted in Fig. 2. Firstly, UBGGset for typical Chinese cities was created for training, validation, and testing of
the deep learning model. Secondly, the novel deep learning model was pre-trained on the UBGGset and tested in
Beijing to compare its performance against state-of-the-art deep learning network models. Finally, the transfer
training was utilized to strengthen the pre-trained model to adapt to diverse landscape characteristics in different
geographic regions and generated UBGG-3m of 36 metropolitan areas in China. Thorough visual inspection and
quantitative accuracy validation were conducted to ensure the reliability and credibility of the UBGG-3m dataset.

UBGG landscapes sample dataset. Accurate and reliable training tags are critical to the accuracy of fine-scale
urban landscape mapping14,20. Current UBGG studies are deficient in standard datasets, so we first created a
large-volume UBGG landscape sample dataset (UBGGset) applicable to urban areas in China. The classification
system includes the UBS, UGS, and UIS landscapes in the city. UBS comprises all water bodies, including rivers,
lakes, and seas, as well as reservoirs and ponds, while UGS is further divided into tree, grass, and farmland. The

Scientific Data | (2024) 11:266 | https://doi.org/10.1038/s41597-023-02844-2 4

www.nature.com/scientificdata/ www.nature.com/scientificdata

Fig. 2 Workflow for generating the Urban Blue-Green-Gray landscape product (UBGG-3m) using high-resolution
Planet satellite images.

others are classified as UIS, including buildings, traffic roads, squares, and other impervious surfaces. In addi-
tion, shaded and bare land is also classified as UIS. UBGGset was constructed with co-registered pairs of 3 m
Planet images and fine-annotated urban landscapes labeled on 1 m Google Earth images. The visual interpreta-
tion process of the UBGGset landscapes was done by the mapping team and further validated by field surveys.
Moreover, UBGGset covers 4 major geographic regions and 15 typical cities (Beijing, Harbin, Changchun, Hefei,
Wuhan, Changsha, Xi’an, Chengdu, Chongqing, Guizhou, Fuzhou, Shenzhen, Hohhot, Lanzhou, and Lhasa),
covering an urban area of about 2,272 km2, which enriches the urban landscape standard datasets and facilitates
the large-scale application of deep network. Examples of UBGGset for six cities are shown in Fig. 3. After that,
50852 training images and 12712 validation images (Length 256 × width 256) were obtained by sliding window
clipping and data enhancement (horizontal, vertical, and diagonal flip).

HRNet-OCR network architecture. The HRNet-OCR network architecture (Fig. 4), constituting the core of the
deep learning model, was designed to tackle the challenges posed by the multi-scale information extraction and
the inadequacy of contextual information in VHR images. Leveraging the High-Resolution Network (HRNet)39
as the backbone network, HRNet-OCR effectively harnessed multi-scale feature learning and exploited four
multi-branch parallel convolutions to generate high-to-low-resolution feature maps40. Meanwhile, multi-scale
information branches are sufficiently linked to enable the seamless flow of information and enhancing semantic
richness and spatial accuracy. Furthermore, this structure can effectively avoid the loss caused by the recovery
of high-resolution features from low-resolution features, thereby preserving the image’s high-resolution features
throughout the process. To overcome the problem of inadequate contextual information, we also integrated
the Object-Contextual Representations (OCR) module31 into model. The OCR module is designed to capture
global context information and integrate it with local features to enhance the model’s ability to recognize and
distinguish objects in VHR images. It also includes a feature fusion component that combines the extracted
features with the original feature map. This process enables the model to integrate both local and global context

Scientific Data | (2024) 11:266 | https://doi.org/10.1038/s41597-023-02844-2 5

www.nature.com/scientificdata/ www.nature.com/scientificdata

Fig. 3 Examples of Urban Blue-Green-Gray sample dataset (UBGGset) for six cities: (a) Beijing, (b) Changchun,
(c) Wuhan, (d) Fuzhou, (e) Chongqing, and (f) Shenzhen. The UBGGset is constructed with co-registered pairs
of 3 m Planet satellite images from © Planet 2020 (left) and fine-annotated urban landscapes labeled (right).

information, thereby improving its capability to recognize objects in complex scenes with multiple objects and
occlusions.
The specific training steps are as follows:
Firstly, we input the train and validation dataset into the HRNet for multi-scale feature learning and then
obtained a coarse segmentation map from the softmax layer.
Secondly, we computed the object regions from the coarse segmentation map from HRNet by aggregating the
representations of all pixels in the Nth landscape object.
∼ x
fN = ∑m Ni i
i∈ τ (1)
Where N is the number of landscape categories, fN represents the N landscape object. xi denotes the pixel i rep-
∼ refers to the normalized degree for pixel i in the Nth landscape object.
resentation. m Ni
Thirdly, we calculated the relationship between pixels and the corresponding landscape object as below:

e N (xi , f N )
wiN =
∑ Nj =1e ( i N )
N x ,f
(2)
Where WiN means the relationship between xi and fN. The transformation function N(x, f) is referenced in
literature31.
Lastly, the final pixel representation zi was obtained by computing the combination of the original representa-
tion xi and the object background representation yi using the transformation function g(·)31.
T
z i = g ([xiT yiT] ) (3)

Scientific Data | (2024) 11:266 | https://doi.org/10.1038/s41597-023-02844-2 6

www.nature.com/scientificdata/ www.nature.com/scientificdata

Fig. 4 The architecture of the HRNet-OCR. The yellow box shows the structure of HRNet39, and the blue box
shows the structure of OCR module31.

 4 
yi = ρ  ∑ wiN δ (f N ) 
N =1  (4)
Where zi is the augmented pixel representation; yi is the object contextual representation.

Transfer learning. To enable the model’s effective performance across large-scale applications, the study
employed a transfer learning technique15. Influenced by natural factors (e.g., vegetation type, spectral diversity
of water bodies, impermeable styles) and external factors (e.g., solar altitude angle, image quality)1, satellite
images collected from different regions could exhibit inconsistent data distribution15. Therefore, a model trained
on one region dataset cannot be applied effectively to images of another region. To overcome this challenge,
transfer learning was employed, where a pre-trained model was used as a starting point for a new task in dif-
ferent geographic regions. Specifically, we first trained a model in the Northen geographic region to obtain a
pre-trained model and further fine-tuned it through adversarial training by adding samples from the next region
to the pre-trained model for parameter initialization and feature extraction (as shown in Fig. 2). This process
was repeated for each geographic region. For large-scale applications, transfer learning can improve computing
efficiency and model generalization compared to starting from scratch on a small sample dataset.

Post-processing. In the post-processing stage of image classification, three essential techniques were imple-
mented to enhance the accuracy of the results. The sliding window prediction method27 was employed to effec-
tively address the issue of insufficient image edge information, mitigating the impact of mosaic traces. Test
enhancement techniques, involving horizontal, vertical, and diagonal flipping, were used to improve classifi-
cation accuracy and reliability by averaging test image enhancements. Lastly, morphological post-processing,
facilitated by the “skimage” package in Python, removed small incorrect patches and filled tiny holes, ensuring
precise and accurate classification results.

Accuracy assessment. To evaluate the accuracy and quality of the proposed UBGG-3m dataset, comprehen-
sive assessments were conducted at the pixel level. The widely used assessment metrics were used to evaluate
classification accuracy for each landscape pixel, including Precision, Recall, overall accuracy (OA), F1-score
(F1), intersection over union (IoU), and frequency weighted intersection over union (FWIoU). The calculation
equations for the metrics are shown in Table 2.

Experimental parameters. The whole experimental process was completed on the High-performance
Computing Platform of Peking University, employing the Pytorch deep learning framework with GPU acceler-
ation from NVIDIA Tesla P100. During the training process, the batch size was 32, and the initial learning rate
was 0.0001. The learning rate was adjusted by simulated annealing to avoid the possibility of the gradient descent
algorithm falling into local minima, with a minimum learning rate of 1e-5. The Adam optimizer was selected for
loss-value optimization with a weight decay factor of 0.001. We set the number of iteration epochs to 120 and
selected the optimal model parameters corresponding to the rounds with the highest accuracy in the training
and validation. The loss function is utilized to calculate the difference between the predicted and true values
and update the network model parameters by error backpropagation. Here, we used the combined loss function
of Soft Cross Entropy Loss (CE) and Dice Loss (DL)5, which can more effectively solve the category imbalance
problem and enhance the model generalization. The calculation formula is as follows:
Loss = wCELossCE + wDLLossDL (5)

Scientific Data | (2024) 11:266 | https://doi.org/10.1038/s41597-023-02844-2 7

www.nature.com/scientificdata/ www.nature.com/scientificdata

Assessment metrics Formula

TP
Precision Precision = TP + FP
TP
Recall Recall = TP + FN
2 × Precision × Recall
F1-Score F1 = Precision + Recall
x
Overall accuracy OA = ∑ ni = 1 Nii
TP
IoU IoU = TP + FP + FN

1 ∑K j = 0 xijxii
FWIoU FWIoU = ∑ iK= 0 K
∑ iK= 0 ∑ K
j = 0 xij ∑ j = 0 xij + ∑ K j = 0 xji − xii

Table 2. Assessment metrics. Where TP, TN, FP and FN represent true positive, true negative, false positive
and false negative; N is the total pixel number of the image. K is the number of categories. xii represents the pixel
number of the category i that was correctly classified. xij represents the pixel number of the category i that are
wrongly divided into category j.

1
LossCE =
N ∑ − [yi ⋅ log (pi ) + (1 − yi ) ⋅ log (1 − pi ) (6)

∑ i ∣pi ∩ yi ∣
LDL = 1 −
∑ i (∣pi ∣ + ∣yi ∣) (7)
where yi is the prediction of urban landscapes of the network. pi is the truth of urban landscapes from label
images. The weights of wCE and wDL are 0.5.

Data Records
The UBGG dataset41 provides easily access and leverage to researchers and analysts, which is stored in the fol-
lowing Zenodo repository (https://doi.org/10.5281/zenodo.8352777). The UBGG dataset consists of two main
components:

1) UBGG-3m: the fine-grained UBGG map of 36 metropolises in China. The UBGG-3m dataset captures
the intricate urban landscape features with remarkable precision, providing a detailed representation at an
impressive 3-meter resolution. The classification maps for all 36 Chinese metropolises were showcased in
Fig. 5. Researchers can delve into the nuances of the UBGG continuum, gaining invaluable insights into
the interplay between the blue, green, and gray elements of urban environments in each metropolis.
2) UBGGset: the large-volume sample dataset to support the UBGG deep learning research. Complementing
the UBGG-3m dataset, UBGGset serves as a large-volume sample dataset specifically tailored to support
and foster UBGG research endeavors. The UBGGset consists of 14,627 sample images (without data aug-
mentation), with dimensions of 256 pixels in length and width, covering an urban area of approximately
2,272 km2.

Technical Validation
Visual and accuracy evaluation on UBGG-3m. To evaluate the accuracy of the product, five cities were
selected for visual and quantitative assessment, namely Beijing, Shenzhen, Harbin, Urumqi, and Lhasa. A total
of six test sample areas were collected in these five cities, covering an area of 43.5 km2. The classification accuracy
was evaluated by comparing the labeled reference maps with the results. The OA for all samples (about 4.83 mil-
lion pixels) was found to be 91.23% (Table 3), indicating promising mapping results. For the different types, UBS
had the highest F1 of 95.15%, followed by UIS at 93.14%. The F1 for trees and grass were 87.54% and 85.97%,
respectively. The quantitative assessment results of different cities were accurate, with OA higher than 91% except
for Lhasa (83.21%), demonstrating the usability and accuracy of the UBGG-3m product.
The visual assessment results of the UBGG-3m were presented in Fig. 6. As the capital of China, Beijing is a
highly urbanized and economically developed city, with comprehensive blue and green infrastructure construc-
tion. The HRNet-OCR model accurately identified UBS ranging from large lakes to small ponds, as well as the
moat surrounding the Forbidden City (Fig. 6a). In addition, the model also effectively captured the sizes, shapes,
locations, and boundaries of UGS, such as individual tree canopies in residential areas, small arborizations in
Peking University, and slender trees on boulevards. Notably, the reconstructed tree geometry was highly consist-
ent with the ground truth data. Moreover, the model was able to successfully distinguish between trees and grass,
highlighting delicate shape contours in areas such as playgrounds of a school and artificial grass on golf courses
(Fig. 6a). The results demonstrated the model’s ability to extract detailed information about the UBGG landscape
in urban areas and to distinguish between different types of greenery with a high level of accuracy. Shenzhen,
located in the southern region of China, is characterized by a higher coverage ratio of UBS and UGS, which is

Scientific Data | (2024) 11:266 | https://doi.org/10.1038/s41597-023-02844-2 8

www.nature.com/scientificdata/ www.nature.com/scientificdata

Fig. 5 Classification maps of Urban Blue-Green-Gray Landscape dataset (UBGG-3m) for 36 Chinese metropolises.

mostly comprised of large reservoirs and parks. The model’s effectiveness in accurately describing the complex
boundary shape of Xikeng Reservoir and identifying trees around commercial and residential buildings, as well
as greenery along roadsides has been demonstrated, as shown in Fig. 6b. Harbin, as a representative city in
northern China, has the largest proportion of farmland within its administrative boundaries. The analysis of
detail maps indicated that building shadows had a certain impact on the accurate extraction of UGS, particularly
in residential areas with tall buildings, as illustrated in Fig. 6c. The presence of building shadows led to discon-
tinuous UGS extraction, and sometimes the obscured areas were classified as UIS, resulting in relatively poor
extraction with F1 of 79.38% and 88.25% for trees and grass, respectively (Table 3). Urumqi, as a representative
of inland cities in the northwest region, has the largest UIS area, and the UBGG-3m product exhibits superior
performance in providing detailed information on UGS and UBS. It is worth noting that despite being geograph-
ically distant from each other, with Harbin located in the far north, Urumqi in the northwest, and Shenzhen
in the south of China, the UBGG classification results for all cities are excellent. This suggests that the model
framework’s performance is unlikely to be affected by geographic location differences, which could be attributed
to the transfer learning strategy that helped the model adapt. Fig. 6e depicts the visualization result of Lhasa, a
representative city in the Qinghai-Tibet region, where the vegetation is primarily comprised of hardy trees and
alpine meadows. The growth of vegetation was affected by the phenological period, and the image of Lhasa city
used in this study was taken on July 24. As the alpine meadows were still in the growing season in July, the UGS

Scientific Data | (2024) 11:266 | https://doi.org/10.1038/s41597-023-02844-2 9

www.nature.com/scientificdata/ www.nature.com/scientificdata

UBS UGS (Tree) UGS (Grass) UIS

City Number of pixels F1 F1 F1 F1 OA mIoU FWIoU
Beijing 1364224 97.53 92.65 89.93 92.61 93.16 87.36 87.29
Harbin 693889 93.70 79.38 88.25 95.10 92.11 80.89 85.45
Shenzhen 1387778 92.01 88.14 70.94 94.52 91.49 77.14 84.51
Urumqi 693889 97.65 88.30 91.84 95.97 94.03 87.90 88.66
Lhasa 693889 92.23 76.40 56.27 82.88 83.21 64.33 71.61
Average 4833669 95.15 87.54 85.97 93.14 91.23 82.78 83.93

Table 3. Quantitative results of accuracy evaluation on UBGG-3m (units: %).

with lower and sparser vegetation cover were more likely to be misclassified, owing to their similarity in appear-
ance to bare ground. Conversely, UGS with higher and denser vegetation cover were more accurately identified.

Comparison with the state-of-the-art deep learning networks. Several state-of-the-art deep learn-
ing networks were selected for performance comparison with HRNet-OCR, including PSPNet42, DeepLabV3+43,
UNet44, HRNet39. The model’s accuracy and loss value records for each round were plotted and shown in Fig. 7.
As depicted in Fig. 7a, the accuracy of all five models increased rapidly with the increase in epochs and then
gradually increased and stabilized after 20 epochs. In terms of training loss, as demonstrated in Fig. 7b, all five
models initially showed a rapid decrease in the first 20 epochs, followed by a more stable decrease. Among the
state-of-the-art deep learning networks, PSPNet exhibited the slowest improvement in classification accuracy and
loss function convergence. Conversely, HRNet outperformed DeepLabv3 + in terms of accuracy improvement
and loss function convergence. Overall, HRNet-OCR demonstrated the most significant training advantage, with
the accuracy reaching 0.989 and the loss reduced to 0.197 after 120 epochs. Although this advantage was not
apparent in the early stage, it showed significant improvement in the later stage compared to HRNet.
The performance evaluation of different models was conducted on a test region covering 1225 ha
(1167 × 1167 pixels) in Haidian District, Beijing, which included Summer Palace Park, Haidian Park, Wanliu
Golf Course, Kunming Lake, and Xiyuan residential area, representing a variety of UBGG landscape fea-
tures (Fig. 7c). The classification results and accuracy assessment of HRNet-OCR and other state-of-the-art
semantic segmentation networks are presented in Fig. 7c and Table 4, respectively. All deep learning meth-
ods demonstrated effective UBS extraction, with F1 above 96.9%. However, the classification of UGS and UIS
was more challenging. PSPNet struggled to handle detailed information, resulting in smooth edges of imper-
vious and trees that were inconsistent with the actual landscape boundaries. DeepLabv3 + still had difficulty
in distinguishing trees and grass, particularly on the golf course lawns, where several solitary tree canopies
were ignored. In comparison, HRNet performed better in classifying UBGG landscapes, particularly in accu-
rately recognizing trees and grasses, with the boundaries of UGS more consistent with actual features, owing
to the high to low resolution feature learning mechanism. Furthermore, the classification accuracy was signifi-
cantly improved by introducing the OCR module based on HRNet. The F1 of UBS, UGS_tree, UGS_grass, and
UIS classified by HRNet-OCR were improved by 0.56%, 1.11%, 1.03%, and 1.95%, respectively, compared to
HRNet. The OA was ranked from large to small: HRNet-OCR (93.16%) >HRNet (91.94%) >UNet (91.05%)
>DeepLabv3 + (91.00%) >PSPNet (89.40%), highlighting the effectiveness and great potential of HRNet-OCR
for high-resolution landscape classification tasks.

Comparison with and without transfer learning in large-scale UBGG landscape classifica-
tion. To develop an ecological understanding of urban systems, the spatial heterogeneity for urban landscapes
from various geographic regions must be addressed for large-scale and fine-grained mapping5,15. Transfer learn-
ing has been demonstrated as a useful tool to address urban landscape heterogeneity and dynamics by a large
body of literature15,45. Our study found that transfer learning can consider the spectral variance of diverse UBS
types, including rivers, lakes, and reservoirs. For example, a large sediment content in the Yellow River causes
a high reflectivity that appears as a blue-green color on a standard false-color image (Fig. 8a), while the Jialing
River shows bright blue color due to its shallow water level, and the Yangtze River has high turbidity showing lake
blue (Fig. 8b). The pre-trained model was unable to fully comprehend this UBS heterogeneity. However, after
transfer learning, the misclassification was much reduced by introducing positive/negative UBS samples and
fine-tuning the pre-trained model with new water features. In addition, the transfer learning cross-geographic
regions method has significant advantages in solving “various UGS in the same spectrum” and “same UGS with
different spectrums”46. For example, crops and urban trees were highly confused due to the same spectral char-
acteristics during the peak growth period of crops in Harbin (Fig. 8d). Similarly, the classification of aquaculture
area in Wuhan also had mixed trees and farmland, manifested by the relatively broken and irregular shape of
farmland patches (Fig. 8e). After adversarial training, the misclassification is much improved, and the edges are
more finely and accurately delineated.
The findings demonstrate that transfer learning can enhance the generalization by efficiently retraining on
the pre-trained model, which is feasible and potentially possible for large-scale, high-resolution UBGG land-
scape mapping. In practical applications, HRNet-OCR can be automatically applied to other cities and achieve
good urban landscape classification by fine-tuning the pre-trained model or even directly using the pre-trained
model. Here, we counted the computational efficiency of the prediction phase. The computational times for

Scientific Data | (2024) 11:266 | https://doi.org/10.1038/s41597-023-02844-2 10

www.nature.com/scientificdata/ www.nature.com/scientificdata

Fig. 6 Classification results of the UBGG-3m in (a) Beijing, (b) Shenzhen, (c) Harbin, (d) Urumqi, and (e) Lhasa.
The small maps at the bottom display detailed classification results of the UBGG landscape in major urban scenes
such as residential areas, schools, parks, etc. (The background images are Planet satellite images from © Planet
2020. The classification results are depicted using colored boundaries, with bright blue representing urban blue
spaces, green indicating trees, yellow representing grass, and orange denoting farmland.).

HRNet-OCR in 36 cities were recorded based on NVIDIA Tesla P100 GPU and Pytorch. Statistically, it only took
about 5 h to generate UBGG-3m covering all 36 metropolitan areas of 50, 411 km2 by transfer learning, which is
effective for timely monitoring and managing dynamic changes in the urban landscape.

Comparison with existing landcover/landscape datasets. Visualization comparisons of UBGG-3m

with existing landcover/landscape datasets are shown in Fig. 9. Additionally, one region of each city was zoomed
in for visual inspection of spatial detail reconstruction. Remarkably, our product demonstrated superior perfor-
mance in terms of visual assessment results, exhibiting excellent landscape classification results. Most of the exist-
ing land cover products, exhibited poor accuracy in reconstructing the UBGG landscape, often misclassifying
blue-green natural land as construction land (Fig. 9). Among the four large-scale land cover products, ESA World
Cover displayed relatively better performance, albeit falling short in accurately depicting the edges of the urban
landscape compared to UBGG-3m. This phenomenon can be attributed to two main factors. Firstly, the diam-
eter of tree crowns typically ranges between 0.5 m and 10 m, and the width of urban rivers and ponds generally
falls between 20 m to 100 m, which can be smaller than one pixel of Sentinel-2 or Landsat21. As a consequence,
the resolution limitation leads to a mixed pixel problem, where scattered UGS and striped UBS may merge with
the surrounding landscape and thus be removed from the pixel47. Secondly, the orientation of these products is
designed for global or national land cover rather than specifically for urban areas37. For example, the Food and
Agriculture Organization (FAO) defines forests as patches greater than 0.5 ha with more than 10% tree canopy
cover, leading to an underestimation of UGS in these products.
Furthermore, our comparative analysis with UTC-2m and UGS-1m demonstrated the superiority of our
UBGG-3m product in accurately capturing urban green space (Figs. 9, 10). The UBGG-3m product based on
higher resolution planet images facilitated more accurate detection of urban tree crowns and finer-grained anal-
ysis of their distribution patterns. In contrast, UTC-2m, due to its lower resolution of Sentinel-2 images, may

Scientific Data | (2024) 11:266 | https://doi.org/10.1038/s41597-023-02844-2 11

www.nature.com/scientificdata/ www.nature.com/scientificdata

Fig. 7 Comparison of (a) accuracy and (b) loss with epoch, and (c) classification results with state-of-the-art
deep learning networks. The classification results include eight panels: (I) Planet satellite images from © Planet
2020, (II) Google images from © Google Earth 2020, (III) Ground truth, (IV) PSPNet, (V) DeepLabv3+, (VI)
UNet, (VII) HRNet, and (VIII) HRNet-OCR.

UBS UGS (Tree) UGS (Grass) UIS

Methods F1 F1 F1 F1 OA mIoU FWIoU
PSPNet 96.98 88.01 85.45 88.28 89.40 81.58 81.05
DeepLabv3+ 97.41 89.84 86.57 90.67 91.00 83.94 83.69
UNet 96.91 90.23 86.60 90.63 91.05 83.86 83.75
HRNet 96.97 91.54 88.90 90.66 91.94 85.36 85.19
HRNet_OCR 97.53 92.65 89.93 92.61 93.16 87.36 87.29

Table 4. Comparison of classification accuracy with state-of-the-art semantic segmentation networks (units: %).

fail to identify small or isolated trees and struggle to distinguish between different types of tree canopies. On
the other hand, UGS-1m, which utilized high-resolution Google imagery, offers a comparable representation of
urban green space when compared to our product. However, the utilization of multispectral information from
Planet imagery allows UBGG-3m to achieve a higher level of discrimination between urban trees, grasslands,
and farmlands, which is not attainable with the other two high-resolution tree products. These comparisons
provide compelling evidence of the superior performance and accuracy of UBGG-3m in capturing the intricate
characteristics of urban landscapes. More importantly, the UBGG-3m product mapped a comprehensive urban
blue-green-gray landscape in human–nature coupled urban systems. It will enable urban planners, researchers,
and policymakers to gain a deeper understanding of the complexities inherent in the urban landscape and facil-
itate more effective management strategies.

Usage Notes
Urban applications. Urban areas occupy only a very small portion of the terrestrial landscape but play a
crucial role in driving environmental change at local, regional, and global scales6,48,49. Although the importance
of urban landscape ecology is increasingly being recognized50, related researches are still limited due to the lack of
large-scale and high-resolution urban landscape maps29,35. With its high resolution and accuracy, the UBGG-3m
product has the potential to provide more precise knowledge of urban landscape and facilitate a deeper under-
standing of the patterns, processes, and implications of urbanization. Here, we briefly describe some research
applications in which our product can be further applied.

(1) Sustainable Urban Planning. UBGG-3m contributes significantly to the development of sustainable urban
planning by providing detailed information on the spatial heterogeneity of landscape types and their
distribution patterns. With the increase in urbanization, the importance of maintaining and enhancing
UGS and UBS has become widely recognized10,51. UBGG-3m enables the identification and quantification
of green and blue infrastructure, which helps in assessing their contributions to urban ecosystems and

Scientific Data | (2024) 11:266 | https://doi.org/10.1038/s41597-023-02844-2 12

www.nature.com/scientificdata/ www.nature.com/scientificdata

Fig. 8 Comparison of classification results before and after transfer learning in urban landscape. (a) Yellow
River Basin in Lanzhou; (b) Yangtze River and Jialing River confluence area in Chongqing; (c) Sand quarries in
Urumqi; (d) Farmland in Harbin; (e) Aquaculture areas in Wuhan.

environmental services. In particular, UBGG-3m allows researchers to analyze the spatial configuration
and pattern of UGS (e.g., tree, grass, and farm), including their connectivity, size, shape, type, and distri-
bution. This information is essential for making informed decisions on urban planning and management,
including land use policies, urban greening, and urban infrastructure development.
(2) Urban thermal environment. Our product contributes to the in-depth study of the urban thermal environ-
ment, where the current understanding of the contributors to the Urban Heat Island (UHI) effect mainly
relies on coarse land cover types due to the lack of high-resolution images6. However, the UHI is more like
an “archipelago” than an “island”52, with local temperature differences as large as those along the urban–ru-
ral gradient. A systematic investigation of the interaction between fine-scale urban landscapes and thermal
environments is still lacking, and UBGG-3m can provide landscapes spatial variation on a fine-scale.
(3) Urban aboveground carbon storage. High-resolution urban landscape products facilitate urban
aboveground carbon storage studies. Numerous studies have proven that UGS have significant carbon sink
potential and provide ecosystem services and livelihood benefits53. However, this service has been largely
underestimated in most studies. For example, an analysis conducted in Beijing showed that carbon stocks
were underestimated by 39% of satellite data from 6 m to 30 m resolution7. Furthermore, according to an
analysis in Leicester, UK54, shifting from 10 m to 250 m resolution remote sensing data resulted in a 76%
underestimation of aboveground carbons stores. Additionally, a survey estimated that more than 1.8 billion
isolated trees in West Africa have carbon stocks up to 22 MgC ha–1, which is far larger than global biomass
mapping23,53. Thus, our product provides essential information on the estimation of urban aboveground
vegetation carbon density with large spatial variability.
(4) Deep learning. This work provides an open high-resolution dataset for urban landscape semantic segmen-
tation studies, which can serve as a huge training pool for high-resolution land cover mapping. Moreover,
Planet images cover a global scale and are freely available, allowing us to develop a robust and transferable
deep network for urban landscape classification using deep learning and transfer learning. At the same
time, our product also promotes more deep learning development models to be applied to urban environ-
mental remote sensing research, driving technological advances in this field and promoting the develop-
ment of urban landscape remote sensing interpretation towards intelligence and automation17.

Apart from the applications discussed above, the UBGG-3m can be combined with big geospatial data and
contribute to other scientific research, such as smart city construction, urban digital twin, sustainability assess-
ment, habitat evaluation, and urban health studies29.

Limitations and future work. This study represents a significant advancement in the production of VHR
urban landscape maps for 36 Chinese metropolises. However, several limitations of the study need to be acknowl-
edged. Firstly, UBGG-3m only covers the 36 cities included in the study, and further work is necessary to extend
this coverage to other cities worldwide. Secondly, the availability of high-resolution images is still limited by
factors such as temporal resolution and cloud cover occlusion. As a result, UBGG-3m only covers the sum-
mer images of 2020-2021. As more high-resolution satellite images become available, future research could be
devoted to landscape classification tasks for more cities and long time series globally. This would provide a more
comprehensive understanding of urban landscape dynamics and aid in developing effective urban planning and
management strategies.

Scientific Data | (2024) 11:266 | https://doi.org/10.1038/s41597-023-02844-2 13

www.nature.com/scientificdata/ www.nature.com/scientificdata

Fig. 9 Visualization comparisons of UBGG-3m with GlobeLand 30 m36, Esri-10m37, SinoLC-1m14, ESA-10m38,
UTC-2m21 and UGS-1m5.

Scientific Data | (2024) 11:266 | https://doi.org/10.1038/s41597-023-02844-2 14

www.nature.com/scientificdata/ www.nature.com/scientificdata

Fig. 10 Visualization comparisons of urban tree extraction between UBGG-3m and high-resolution urban
green space dataset in Beijing. (a) Planet satellite images from © Planet 2020; (b) Google Images from ©
Google Earth 2020; (c) Comparison of UBGG-3m and Urban Tree Cover-2m (UTC-2m)21; (d) Comparison
of UBGG-3m and Urban Green Space-1m (UGS-1m)5. The green region represents the agreement between
UBGG-3m and the other products in identifying urban trees. The yellow region represents the urban trees
underestimated by other products compared to UBGG-3m, while the blue region represents the overestimated
area by other products compared to UBGG-3m.

Code availability
The programs used to generate the dataset were ENVI (5.3), ESRI ArcGIS (10.6) and Pytorch deep learning
framework. All used codes to generate the dataset are available in the following GitHub (https://github.com/
Zhiyu-Xu/Fine-grained-urban-blue-green-gray-landscape-dataset-for-36-Chinese-cities).

Received: 24 July 2023; Accepted: 11 December 2023;

Published: xx xx xxxx

References
1. Cadenasso, M. L., Pickett, S. T. & Schwarz, K. Spatial heterogeneity in urban ecosystems: reconceptualizing land cover and a
framework for classification. Frontiers in Ecology and the Environment. 5, 80–88, https://doi.org/10.1890/1540-9295(2007)5[80:SHI
UER]2.0.CO;2 (2007).
2. Brelsford, C., Lobo, J., Hand, J. & Bettencourt, L. M. A. Heterogeneity and scale of sustainable development in cities. Proceedings of
the National Academy of Sciences. 114, 8963–8968, https://doi.org/10.1073/pnas.1606033114 (2017).
3. Gunawardena, K. R., Wells, M. J. & Kershaw, T. Utilising green and bluespace to mitigate urban heat island intensity. Science of The
Total Environment. 584-585, 1040–1055, https://doi.org/10.1016/j.scitotenv.2017.01.158 (2017).
4. Kuang, W., Zhang, S., Li, X. & Lu, D. A 30-m resolution dataset of China’s urban impervious surface area and green space,
2000–2018. Earth Syst. Sci. Data. 13, 63–82, https://doi.org/10.5194/essd-13-63-2021 (2021).
5. Shi, Q., Liu, M., Marinoni, A. & Liu, X. UGS-1m: fine-grained urban green space mapping of 31 major cities in China based on the
deep learning framework. Earth Syst. Sci. Data. 15, 555–577, https://doi.org/10.5194/essd-15-555-2023 (2023).
6. Zhou, D., Zhao, S., Liu, S., Zhang, L. & Zhu, C. Surface urban heat island in China’s 32 major cities: Spatial patterns and drivers.
Remote Sensing of Environment. 152, 51–61, https://doi.org/10.1016/j.rse.2014.05.017 (2014).
7. Sun, Y., Xie, S. & Zhao, S. Valuing urban green spaces in mitigating climate change: A city‐wide estimate of aboveground carbon
stored in urban green spaces of China’s Capital. Global Change Biology. 25, 1717–1732, https://doi.org/10.1111/gcb.14566 (2019).
8. Veerkamp, C. J. et al. A review of studies assessing ecosystem services provided by urban green and blue infrastructure. Ecosystem
Services. 52, 101367, https://doi.org/10.1016/j.ecoser.2021.101367 (2021).
9. Yang, L., Zhao, S. & Liu, S. A global analysis of urbanization effects on amphibian richness: Patterns and drivers. Global
Environmental Change. 73, 102476, https://doi.org/10.1016/j.gloenvcha.2022.102476 (2022).
10. Andersson, E. et al. Enabling Green and Blue Infrastructure to Improve Contributions to Human Well-Being and Equity in Urban
Systems. BioScience. 69, 566–574, https://doi.org/10.1093/biosci/biz058 (2019).
11. Feng, M. & Li, X. Land cover mapping toward finer scales. Science Bulletin. 65, 1604–1606, https://doi.org/10.1016/j.scib.2020.06.014
(2020).
12. Qian, Y. et al. Integrating structure and function: mapping the hierarchical spatial heterogeneity of urban landscapes. Ecological
Processes. 9, 59, https://doi.org/10.1186/s13717-020-00266-1 (2020).

Scientific Data | (2024) 11:266 | https://doi.org/10.1038/s41597-023-02844-2 15

www.nature.com/scientificdata/ www.nature.com/scientificdata

13. Pickett, S. T. A. et al. Dynamic heterogeneity: a framework to promote ecological integration and hypothesis generation in urban
systems. Urban Ecosystems. 20, 1–14, https://doi.org/10.1007/s11252-016-0574-9 (2017).
14. Li, Z. et al. SinoLC-1: the first 1 m resolution national-scale land-cover map of China created with a deep learning framework and
open-access data. Earth Syst. Sci. Data. 15, 4749-4780, https://doi.org/10.5194/essd-15-4749-2023 (2023).
15. Tong, X.-Y. et al. Land-cover classification with high-resolution remote sensing images using transferable deep models. Remote
Sensing of Environment. 237, 111322, https://doi.org/10.1016/j.rse.2019.111322 (2020).
16. Zhu, X. X. et al. Deep Learning in Remote Sensing: A Comprehensive Review and List of Resources. IEEE Geoscience and Remote
Sensing Magazine. 5, 8–36, https://doi.org/10.1109/MGRS.2017.2762307 (2017).
17. Li, J., Huang, X. & Gong, J. Deep neural network for remote-sensing image interpretation: status and perspectives. National Science
Review. 6, 1082–1086, https://doi.org/10.1093/nsr/nwz058 (2019).
18. LeCun, Y., Bengio, Y. & Hinton, G. Deep learning. Nature. 521, 436–444, https://doi.org/10.1038/nature14539 (2015).
19. Yuan, Q. et al. Deep learning in environmental remote sensing: Achievements and challenges. Remote Sensing of Environment. 241,
111716, https://doi.org/10.1016/j.rse.2020.111716 (2020).
20. Li, Z., Zhang, X. & Xiao, P. Spectral index-driven FCN model training for water extraction from multispectral imagery. ISPRS
Journal of Photogrammetry and Remote Sensing. 192, 344–360, https://doi.org/10.1016/j.isprsjprs.2022.08.019 (2022).
21. He, D., Shi, Q., Liu, X., Zhong, Y. & Zhang, L. Generating 2m fine-scale urban tree cover product over 34 metropolises in China
based on deep context-aware sub-pixel mapping network. International Journal of Applied Earth Observation and Geoinformation.
106, 102667, https://doi.org/10.1016/j.jag.2021.102667 (2022).
22. Xu, Z., Zhou, Y., Wang, S., Wang, L. & Wang, Z. U-Net for urban green space classification in Gaofen-2 remote sensing images.
Journal of Image and Graphics. 26, 0700–0713, https://doi.org/10.11834/jig.200052 (2021).
23. Brandt, M. et al. An unexpectedly large count of trees in the West African Sahara and Sahel. Nature. 587, 78–82, https://doi.
org/10.1038/s41586-020-2824-5 (2020).
24. Wang, Y., Li, Z., Zeng, C., Xia, G. S. & Shen, H. An Urban Water Extraction Method Combining Deep Learning and Google Earth
Engine. IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing. 13, 769–782, https://doi.org/10.1109/
JSTARS.2020.2971783 (2020).
25. Chen, F. et al. Open water detection in urban environments using high spatial resolution remote sensing imagery. Remote Sensing of
Environment. 242, 111706, https://doi.org/10.1016/j.rse.2020.111706 (2020).
26. Chen, Z. et al. Corse-to-fine road extraction based on local Dirichlet mixture models and multiscale-high-order deep learning. IEEE
Transactions on Intelligent Transportation Systems. 21, 4283–4293, https://doi.org/10.1109/TITS.2019.2939536 (2019).
27. Wang, Z., Zhou, Y., Wang, S., Wang, F. & Xu, Z. House building extraction from high resolution remote sensing image based on
IEU-Net. Journal of Remote Sensing. 025, 2245–2254, https://doi.org/10.11834/jrs.20210042 (2021).
28. Guo, H., Du, B., Zhang, L. & Su, X. A coarse-to-fine boundary refinement network for building footprint extraction from remote
sensing imagery. ISPRS Journal of Photogrammetry and Remote Sensing. 183, 240–252, https://doi.org/10.1016/j.isprsjprs.2021.11.005
(2022).
29. Huang, X. et al. High-resolution urban land-cover mapping and landscape analysis of the 42 major cities in China using ZY-3
satellite images. Science Bulletin. 65, 1039–1048, https://doi.org/10.1016/j.scib.2020.03.003 (2020).
30. Zhou, W., Cadenasso, M. L., Schwarz, K. & Pickett, S. T. A. Quantifying Spatial Heterogeneity in Urban Landscapes: Integrating
Visual Interpretation and Object-Based Classification. Remote Sensing. 6, 3369–3386, https://doi.org/10.3390/rs6043369 (2014).
31. Yuan, Y., Chen, X. & Wang, J. in Computer Vision – ECCV 2020. (eds Vedaldi, A., Bischof, H., Brox, T. & Frahm, J.-M.) 173–190
(Springer International Publishing), https://doi.org/10.1007/978-3-030-58539-6_11 (2020).
32. Zhao, S. et al. Rates and patterns of urban expansion in China’s 32 major cities over the past three decades. Landscape Ecology. 30,
1541–1559, https://doi.org/10.1007/s10980-015-0211-7 (2015).
33. Du, S., Du, S., Liu, B. & Zhang, X. Mapping large-scale and fine-grained urban functional zones from VHR images using a multi-
scale semantic segmentation network and object based approach. Remote Sensing of Environment. 261, 112480, https://doi.
org/10.1016/j.rse.2021.112480 (2021).
34. Liang, L. et al. Long-term spatial and temporal variations of vegetative drought based on vegetation condition index in China.
Ecosphere. 8, e01919, https://doi.org/10.1002/ecs2.1919 (2017).
35. Zhu, Z. et al. Understanding an urbanizing planet: Strategic directions for remote sensing. Remote Sensing of Environment. 228,
164–182, https://doi.org/10.1016/j.rse.2019.04.020 (2019).
36. Jun, C., Ban, Y. & Li, S. Open access to Earth land-cover map. Nature. 514, 434–434, https://doi.org/10.1038/514434c (2014).
37. Karra, K. et al. in 2021 IEEE International Geoscience and Remote Sensing Symposium IGARSS. 4704–4707, https://doi.org/10.1109/
IGARSS47720.2021.9553499 (2021).
38. Zanaga, D. et al. ESA WorldCover 10 m 2020 v100. Zenodo https://doi.org/10.5281/zenodo.5571936 (2021).
39. Wang, J. et al. Deep High-Resolution Representation Learning for Visual Recognition. IEEE Transactions on Pattern Analysis and
Machine Intelligence. 43, 3349–3364, https://doi.org/10.1109/TPAMI.2020.2983686 (2021).
40. Xu, Z. & Zhao, S. Scale dependence of urban green space cooling efficiency: A case study in Beijing metropolitan area. Science of The
Total Environment. 898, 165563, https://doi.org/10.1016/j.scitotenv.2023.165563 (2023).
41. Xu, Z. & Zhao, S. UBGG-3m: Fine-grained urban blue-green-gray landscape dataset for 36 Chinese cities based on deep learning
network. Zenodo https://doi.org/10.5281/zenodo.8352777 (2023).
42. Zhao, H. et al. in 30th IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR). 6230–6239 (2017), https://doi.
org/10.1109/cvpr.2017.660 (2017).
43. Chen, L.-C., Zhu, Y., Papandreou, G., Schroff, F. & Adam, H. in 15th European Conference on Computer Vision (ECCV). 833–851
(2018), https://doi.org/10.1007/978-3-030-01234-2_49 (2018).
44. Ronneberger, O., Fischer, P. & Brox, T. in 18th International Conference on Medical Image Computing and Computer-Assisted
Intervention (MICCAI). 234-241 (2015), https://doi.org/10.1007/978-3-319-24574-4_28 (2015).
45. Wurm, M., Stark, T., Zhu, X. X., Weigand, M. & Taubenböck, H. Semantic segmentation of slums in satellite images using transfer
learning on fully convolutional neural networks. ISPRS Journal of Photogrammetry and Remote Sensing. 150, 59–69, https://doi.
org/10.1016/j.isprsjprs.2019.02.006 (2019).
46. Xu, Z. et al. A novel intelligent classification method for urban green space based on high-resolution remote sensing images. Remote
sensing. 12, 3845, https://doi.org/10.3390/rs12223845 (2020).
47. Qian, Y., Zhou, W., Yu, W. & Pickett, S. T. A. Quantifying spatiotemporal pattern of urban greenspace: new insights from high
resolution data. Landscape Ecology. 30, 1165–1173, https://doi.org/10.1007/s10980-015-0195-3 (2015).
48. Grimm, N. B. et al. Global Change and the Ecology of Cities. Science. 319, 756–760, https://doi.org/10.1126/science.1150195 (2008).
49. Zhao, S., Liu, S. & Zhou, D. Prevalent vegetation growth enhancement in urban environment. Proceedings of the National Academy
of Sciences. 113, 6313–6318, https://doi.org/10.1073/pnas.1602312113 (2016).
50. Wu, J., He, C., Huang, G. & Yu, D. in Landscape Ecology for Sustainable Environment and Culture (eds Fu, B. & Jones, K. B.) 37–53,
https://doi.org/10.1007/978-94-007-6530-6_3 (Springer Netherlands.Press, 2013).
51. Wong, N. H., Tan, C. L., Kolokotsa, D. D. & Takebayashi, H. Greenery as a mitigation and adaptation strategy to urban heat. Nature
Reviews Earth & Environment. 2, 166–181, https://doi.org/10.1038/s43017-020-00129-5 (2021).

Scientific Data | (2024) 11:266 | https://doi.org/10.1038/s41597-023-02844-2 16

www.nature.com/scientificdata/ www.nature.com/scientificdata

52. Ziter, C. D., Pedersen, E. J., Kucharik, C. J. & Turner, M. G. Scale-dependent interactions between tree canopy cover and impervious
surfaces reduce daytime urban heat during summer. Proceedings of the National Academy of Sciences. 116, 7575–7580, https://doi.
org/10.1073/pnas.1817561116 (2019).
53. Skole, D. L., Mbow, C., Mugabowindekwe, M., Brandt, M. & Samek, J. H. Trees outside of forests as natural climate solutions. Nature
Climate Change. 11, 1013–1016, https://doi.org/10.1038/s41558-021-01230-3 (2021).
54. Davies, Z. G., Dallimer, M., Edmondson, J. L., Leake, J. R. & Gaston, K. J. Identifying potential sources of variability between
vegetation carbon storage estimates for urban areas. Environmental Pollution. 183, 133–142, https://doi.org/10.1016/j.
envpol.2013.06.005 (2013).

Acknowledgements
This study was supported by the National Key R&D Plan of China Grant (No.2023YFF1304602) and the National
Natural Science Foundation of China Grant (No. 42071120). Computational work was supported by resources
provided by the High-performance Computing Platform of Peking University.

Author contributions
Zhiyu Xu: Methodology, Software, Formal analysis, Data curation, Writing – original draft, Writing – review &
editing, Visualization. Shuqing Zhao: Conceptualization, Methodology, Formal analysis, Writing – original draft,
Writing – review & editing, Supervision, Project administration, Funding acquisition.

Competing interests
The authors declare no competing interests.

Additional information
Correspondence and requests for materials should be addressed to S.Z.
Reprints and permissions information is available at www.nature.com/reprints.
Publisher’s note Springer Nature remains neutral with regard to jurisdictional claims in published maps and
institutional affiliations.
Open Access This article is licensed under a Creative Commons Attribution 4.0 International
License, which permits use, sharing, adaptation, distribution and reproduction in any medium or
format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Cre-
ative Commons licence, and indicate if changes were made. The images or other third party material in this
article are included in the article’s Creative Commons licence, unless indicated otherwise in a credit line to the
material. If material is not included in the article’s Creative Commons licence and your intended use is not per-
mitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the
copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by/4.0/.

© The Author(s) 2024

Scientific Data | (2024) 11:266 | https://doi.org/10.1038/s41597-023-02844-2 17