Datasets for simulating heat and CO2 fluxes in Beijing using SUEWS V2020b

Data, model runs and codes used in manuscript "Simulating heat and CO2 fluxes in Beijing using SUEWS V2020b: sensitivity to vegetation phenology and maximum conductance". The source code for Surface Urban Energy and Water balance Scheme (SUEWS) V2020b is also included. The folders are: 1. Observations The units for CO2 flux (Fc) , latent heat flux (QE) and sensible heat flux (QH) are μmol m-2 s-1, W m-2, and W m-2, respectively, unless otherwise stated. 1.1. Observation datasets 1.1.1. radiation_flux_2010-2011.csv is radiation flux observations from May 2010 to July 2011 at 140 m on the IAP tower, Beijing. 1.1.2. turbulent_flux_before_QC_2016.csv is turbulent flux observations before quality control (QC) obtained with eddy covariance (EC) technique for the year 2016 at 47 m on the IAP tower. 1.1.3. turbulent_flux_after_QC_2016.csv is turbulent flux observations after quality control (QC). 1.1.4. Fc_mean_diurnal_by_season.csv is the mean diurnal seasonal cycle of Fc observations. 1.1.5. Fc_gapfilled_with_MDC.csv is Fc time series for the year 2016 gap-filled with Mean Diurnal Cycle (MDC) method. 1.1.6. meteorological_observations is a folder including meteorological observations at different heights measured at the IAP tower for the years 2010-2012. 1.2 Codes for observations processing 1.2.1. turbulent_flux_QC.py is for turbulent flux QC covering wind direction filtering, stationarity test, friction velocity filtering, and nighttime filtering. 1.2.2. mean_diurnal_cycle_of_flux_observations.py is for obtaining the mean diurnal cycle of flux observations. 1.2.3. Fc_gap_filling.py is for filling the gaps in Fc time series with MDC method. 2. LAI This folder includes a project aiming to obtain the parameters for the leaf area index (LAI) model adopted by SUEWS. In this LAI model, the LAI is related to air temperature only. Incorporating hourly air temperature observations and the 'real' LAI (e.g. observed LAI, remotely sensed LAI), the optimization method Covariance Matrix Adaptation Evolutionary Strategies (CMA-ES) will derive the optimized parameters for the LAI model. For more details, please read the description text file (read_me.txt) under this folder. 2.1. read_me.txt. 2.2. source_data is a folder including the input for CMA-ES optimization. The input files of Beijing are included, allowing a quick run to reproduce the results demonstrated in the manuscript. 2.3. result_data is a folder including the output for CMA-ES optimization. 2.4. The others are Python scripts regarding CMA-ES, including the LAI model, data de-spiking, interpolation, CMA-ES training, and result visualization. 3. ModelRuns This folder includes SUEWS source code, four model runs conducted to test the sensitivity to vegetation-related parameters (i.e., maximum conductance, LAI model parameters), and four model runs to test sensitivity to the radius of model domain. 3.1. SUEWS_SourceCode is a folder including SUEWS V2020b source Fortran codes. For detailed description, readers are referred to SUEWS webpage (https://suews.readthedocs.io/en/latest/). 3.2. Input is a folder including the necessary input files shared by each of the model runs. Please copy all the files to a model run folder (e.g. VegetationSensitivity\case-base\Input\) before starting a SUEWS run. 3.3. VegetationSensitivity is a folder including four SUEWS runs to test the model sensitivity to vegetation-related parameters: case-base is the control run, case-gs is a model run where maximum conductance of vegetation has been modified to more site-specific values, case-LAI is a model run where LAI parameters have been optimized with CMA-ES, and case-gs_LAI is a model run where both maximum conductance and LAI parameters have been modified. Note that all the model runs have selected 1000 m as the radius of the model domain. To conduct a quick model run to reproduce the output, enter the command "./SUEWS_V2020b" in the model run folder (e.g. VegetationSensitivity\case-base\) under a Linux environment (e.g., Ubuntu). 3.4. RadiusSensitivity is a folder including four SUEWS runs to test the model sensitivity to the size of model domain, covering circles with a 500 m, 750 m, 1000 m, 1500 m radius, respectively. Note that except the size of model domain, the parameters are the same as model run VegetationSensitivity\case-gs_LAI. 4. Fig&Statistics This folder includes Python scripts and data to reproduce a portion of the figures and statistics in the manuscript quickly. They are sorted into the following sub-folders: 4.1 validation_WFDE5, 4.2 evaluation_radiation, 4.3 evaluation_turbulent, 4.4 accumulated_Fc, 4.5 Fc_component_MDC