DongchenZ.github.io/index.html at main · DongchenZ/DongchenZ.github.io · GitHub

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
<!DOCTYPE html>
<html lang="en">
<head>
    <meta charset="UTF-8">
    <meta name="viewport" content="width=device-width, initial-scale=1.0">
    <title>Dongchen Zhang - Portfolio</title>
    <style>
        body {
            font-family: 'Segoe UI', Tahoma, Geneva, Verdana, sans-serif;
            line-height: 1.6;
            color: #333;
            max-width: 900px;
            margin: 0 auto;
            padding: 20px;
            background-color: #f9f9f9;
        }
        header {
            text-align: center;
            padding: 40px 0;
            background-color: #2c3e50;
            color: white;
            border-radius: 8px;
            margin-bottom: 30px;
        }
        header h1 {
            margin: 0;
            font-size: 2.5em;
        }
        header p {
            margin: 10px 0 0;
            font-size: 1.1em;
            color: #ecf0f1;
        }
        header a {
            color: #3498db;
            text-decoration: none;
        }
        section {
            background: white;
            padding: 25px;
            margin-bottom: 25px;
            border-radius: 8px;
            box-shadow: 0 2px 5px rgba(0,0,0,0.05);
        }
        h2 {
            color: #2c3e50;
            border-bottom: 2px solid #3498db;
            padding-bottom: 10px;
            margin-top: 0;
        }
        h3 {
            color: #2980b9;
            margin-bottom: 5px;
        }
        .date {
            font-style: italic;
            color: #7f8c8d;
            font-size: 0.9em;
            margin-bottom: 15px;
            display: block;
        }
        ul {
            padding-left: 20px;
        }
        li {
            margin-bottom: 10px;
        }
        .publication {
            margin-bottom: 15px;
        }
    </style>
</head>
<body>

    <header>
        <h1>Dongchen Zhang</h1>
        <p>Boston, MA 02155 | zhangdc@bu.edu | (617)-320-9561</p>
        <p><a href="https://github.com/DongchenZ" target="_blank">https://github.com/DongchenZ</a></p>
    </header>

    <section id="education">
        <h2>Education</h2>

        <h3>Boston University</h3>
        <span class="date">09/2021 - Present</span>
        <p>Ph.D. in Earth and Environment | GPA: 3.97/4.0</p>

        <h3>Boston University</h3>
        <span class="date">09/2019 - 01/2020</span>
        <p>MA in Remote Sensing and Geosciences | GPA: 3.63/4.0</p>

        <h3>China University of Geosciences, Wuhan</h3>
        <span class="date">09/2014 - 06/2018</span>
        <p>Bachelor of Geo-information Science and Technology | GPA: 3.53/4.0; Specialized GPA: 3.72</p>
    </section>

    <section id="experience">
        <h2>Research Experience</h2>

        <h3>Graduate Research Assistant | Ecological Forecasting Lab | Boston University</h3>
        <span class="date">09/2021 - Present</span>

        <h4>Project 1: GEDI biomass assimilation across North America (NA)</h4>
        <span class="date">02/2026 - Present</span>
        <ul>
            <li>Developed an automated HPC parallel pipeline to retrieve 4.75 TB of GEDI data, reducing preprocessing latency by >1000X.</li>
            <li>Designed and implemented a novel aggregation algorithm for GEDI footprint biomass, enabling seamless spatiotemporal data fusion at 1 km² resolution across North America.</li>
            <li>Leveraged machine learning to build a cross-product bias correction model, ensuring data integrity and consistency across multiple biomass variables.</li>
            <li>Assimilated GEDI biomass time series into the North American carbon cycling framework, reducing predictive uncertainty by 70% and boosting R² by 0.38 in tropical regions (e.g., Mexico and Central America).</li>
        </ul>

        <h4>Project 2: Estimate North American (NA) carbon cycling using data assimilation</h4>
        <span class="date">09/2024 - 03/2026</span>
        <ul>
            <li>Formulated a spatial clustering algorithm to optimize site selection across diverse eco-climatic gradients, covering the entire North American continent.</li>
            <li>Implemented a Bayesian Data Assimilation infrastructure, ensuring seamless uncertainty propagation throughout the data-model fusion pipeline.</li>
            <li>Improved model accuracy by developing a machine-learning bias-correction module, reduced maximum RMSE by up to 79%.</li>
            <li>Upscaled predictive modeling to 21 million pixels at 1 km² resolution, utilizing a multi-variate feature set (14 covariates) to estimate NA carbon budgets.</li>
            <li>Optimized HPC parallel processing to handle ~10 TB of spatiotemporal data, achieving high-frequency (3-hour interval) output for a 12-year period.</li>
            <li>Provided the most comprehensive and up-to-date North American carbon cycling estimates in the field, along with uncertainty and spatial covariances.</li>
        </ul>

        <h4>Project 3: Uncertainty propagation and analysis in carbon cycling data assimilation</h4>
        <span class="date">04/2023 - 08/2025</span>
        <ul>
            <li>Created a robust Bayesian infrastructure for data-model fusion, providing a scalable solution for end-to-end uncertainty propagation and analysis.</li>
            <li>Reduced model uncertainty by propagating uncertainties through process chains, reducing standard deviation by up to 62% in multi-variable carbon estimates.</li>
            <li>Processed and analyzed decadal (2012-2021) spatiotemporal data for 39 CONUS NEON sites, generating 0.7 million high-resolution (3-hourly) carbon and water budget records.</li>
            <li>Integrated spatial covariance structures into the assimilation pipeline, leveraging regional correlations to reduce joint uncertainty by 30% on average.</li>
        </ul>

        <h4>Project 4: Estimate daily canopy biochemical traits using remote sensing</h4>
        <span class="date">05/2021 - 03/2023</span>
        <ul>
            <li>Architected a multi-sensor data fusion framework by coupling Bayesian Markov Chain Monte Carlo (MCMC) with Radiative Transfer Models (RTM), integrating MODIS, Landsat, and Sentinel-2 observations.</li>
            <li>Developed a robust cross-sensor cloud masking algorithm, ensuring spatiotemporal data consistency and seamless integration across heterogeneous satellite platforms.</li>
            <li>Engineered a Bayesian state-space time-series model to derive daily canopy biochemical trait estimates, effectively interpolating sparse satellite revisits with 650% higher temporal frequency.</li>
            <li>Modeled complex error structures, including temporal variability and measurement noise, to provide 95% confidence intervals for all estimated traits.</li>
            <li>Validated retrieval accuracy against 40 extensive ground-truth sites, demonstrating high model accuracy with an R² of 0.6 across diverse vegetation types.</li>
        </ul>
    </section>

    <section id="publications">
        <h2>Articles</h2>
        <div class="publication">
            <p><strong>Zhang, D.</strong>, Li, Q., Helgeson, A., Serbin, S. P., & Dietze, M. (2026). Harmonizing terrestrial carbon cycle observations over CONUS NEON sites: Assessing the information contributions of multiple data constraints. <em>Global Change Biology, 32(2)</em>, e70761.</p>
        </div>
        <div class="publication">
            <p><strong>Zhang, D.</strong>, & Dietze, M. (2023). Towards uninterrupted canopy-trait time-series: A Bayesian radiative transfer model inversion using multi-sourced satellite observations. <em>Remote Sensing of Environment, 287</em>, 113475.</p>
        </div>
        <div class="publication">
            <p><strong>Zhang, D.</strong>, Li, Q., Zuo, Z., Ramachandran, S., Serbin, S., Webb, C., & Dietze, M. (2026). Mapping the North American Terrestrial Carbon Cycle: A Process-based Reanalysis Using State Data Assimilation (SDA). <em>Remote Sensing of Environment (Under Review)</em>.</p>
        </div>
        <div class="publication">
            <p>Zhenpeng Zuo, Hangkai You, Katie Glodzik, <strong>Dongchen Zhang</strong>, Yupan Zhang, Yuqi Duan, Ranga Myneni. Mapping potential forest canopy top height over the Eastern United States for aid in restoration planning. <em>Agricultural and Forest Meteorology (Under Review)</em>.</p>
        </div>
        <div class="publication">
            <p>Qianyu Li, <strong>Dongchen Zhang</strong>, Alexis Helgeson, Michael Dietze, Shawn P. Serbin. Soil carbon assimilation effectively constrains carbon-cycle model forecasting. <em>Biogeoscience (Under Review)</em>.</p>
        </div>
        <div class="publication">
            <p><strong>Zhang, D.</strong>, Dietze, M. (2025). North American terrestrial carbon cycling data assimilation of GEDI Aboveground biomass time series. <em>(In preparation)</em>.</p>
        </div>
    </section>

    <section id="skills">
        <h2>Skills</h2>
        <ul>
            <li><strong>Programming:</strong> Python (NumPy, Pandas, Scikit-learn), R (tidyverse, sf, ggplot), C/C++, Java, MATLAB, SQL.</li>
            <li><strong>Earth Observation & GIS:</strong> Google Earth Engine (GEE), LIDAR (GEDI), Multi-sensor Data Fusion (MODIS, Landsat, Sentinel), Radiative Transfer Modeling (RTM/PROSAIL).</li>
            <li><strong>Data Science & Modeling:</strong> Bayesian Inference (MCMC, State-space Time Series), Data Assimilation, Machine Learning, Geospatial Statistics, Uncertainty Quantification.</li>
            <li><strong>Infrastructure & Tools:</strong> HPC (SLURM/PBS), Linux/Unix CLI, GitHub Version Control, Automated Pipelines.</li>
        </ul>
    </section>

</body>
</html>