Thirty-minute time series of observed fluxes and scalars from the Canopy Horizontal Array Turbulence Study (CHATS) field project for land surface model evaluation

Information for “Thirty-minute time series of observed fluxes and scalars from the Canopy Horizontal Array Turbulence Study (CHATS) field project for land surface model evaluation,” by Sean P. Burns and Gordon B. Bonan https://doi.org/10.5281/zenodo.17426255 Data File Description [email protected] Date: December 5, 2025 The NSF National Center for Atmospheric Research (NSF NCAR) Canopy Horizontal Array Turbulence Study (CHATS) field project took place in a California walnut orchard in 2007 from mid-March to mid-June (Patton et al., 2011). This document contains additional information about the 30-min CHATS dataset archived in the NSF NCAR Geoscience Data Exchange— Integrated Research Data Commons (GDEX) database, which uses zenodo. The 30-min CHATS data files were created from the NSF NCAR EOL 5-min statistics netCDF data files (Horst, 2020) for comparing the CHATS observations with the Community Land Model (CLM) multilayer canopy model (CLM-ml v2; Bonan et al., 2026). The calculations and other data details of the 30-min fluxes are described in Bonan et al. (2026) which is partially replicated below in Sect. 4. Below, we highlight the specific details about the 30-min dataset as well as which variable is within each column of the CSV data files. 1. Data Files Details We use 5-min mean and covariance statistics from version 2 of the NSF NCAR Earth Observing Laboratory (EOL) high-rate CHATS data (Horst and Oncley, 2019). Version 2 includes a sonic anemometer tilt-correction (Wilczak et al., 2001) and shadow-correction for the CSAT3 transducers (Horst et al., 2015). The EOL CHATS 5-min covariance/statistics netCDF data files have been quality-controlled (i.e., incorrect data removed, despiked, etc.) with all the appropriate calibrations applied (Horst, 2020). Additional information about the sensors and data quality-control is also on EOL CHATS website (CHATS project webpage, 2007). The 5-min data are distributed by EOL in netCDF files with each file covering a 24 hour period (i.e., 288 fiveminute samples in each netCDF file) using a file-naming scheme “isff_2007MMDD.nc”, where MM corresponds to the month and DD to the day in UTC. 1 From these 5-min netCDF data files, we extracted a subset of the entire CHATS data set, that focused on radiation, wind, turbulence, fluxes, and scalar profiles from the CHATS tall tower. These vertical profiles of atmospheric quantities were from multiple levels both above and within the orchard canopy. We also included 30-min averages of the soil heat flux. In all, 151 different variables are included in the the 30-min CHATS CSV data files as specified in Sect. 3. For April and May 2007 the CSV data files have a size of around 2 Mbyte each: 1955404 Aug 25 10:21 chats_30min_data_2007_04_ver_250801.csv 2043695 Aug 1 23:58 chats_30min_data_2007_05_ver_250801.csv The time stamp within the data files are in UTC (PST=Pacific Standard Time) and the following time periods are covered by each CSV data file: chats_30min_data_2007_04_ver_250801.csv: PST Time Period: 2007 03/31 16:00:00 - 2007 04/30 16:00:00, DOY 90.667-120.667 (PST) UTC Time Period: 2007 04/01 00:00:00 - 2007 05/01 00:00:00, DOY 91.000-121.000 (UTC) chats_30min_data_2007_05_ver_250801.csv: PST Time Period: 2007 04/30 16:00:00 - 2007 05/31 16:00:00, DOY 120.667-151.667 (PST) UTC Time Period: 2007 05/01 00:00:00 - 2007 06/01 00:00:00, DOY 121.000-152.000 (UTC) 2. Notes about the 30-min CSV Data Files Additional notes about the 30-min CSV data files are as follows: •Missing data are set to 1e+36 •The 30-min averaging period is between the top of the hour and 30-min past the hour. The UTC time stamp within the CSV data file represents the middle of the 30-min averaging time period. As an example, here are first few samples in chats_30min_data_2007_05_ver_250801.csv: 2007, 5, 1, 0, 15, 0, 121.0104, 502.7423, 83.94901, 342.9013, 462.9205, 473.6891, 9.235837, etc, 2007, 5, 1, 0, 45, 0, 121.0312, 393.3702, 68.89728, 338.628, 457.49, 250.3904, -1.68107, etc, 2007, 5, 1, 1, 15, 0, 121.0521, 293.1889, 54.03268, 338.7015, 450.8514, 54.20478, -6.192818, etc. •Soil heat flux plate data are with and without the Philip correction (Philip, 1961). •The u,v,w wind components are in a planar-fit geographic (ie, east/west and north/south) coordinate system. The "global attributes" in the EOL-generated 5-min netCDF data file, describe this as follows: // global attributes: :history = "Created: 2015-10-25 06:57:46 +0000\n", :NIDAS_version = "v1.2-13" ; :calibration_file_path = "/net/isf/isff/projects/CHATS/ISFF/cal_files/QC,version=1561:/net/isf/isff/cal_files,version=550" ; :dataset = "qc_geo_tiltcor_shadowcor" ; :dataset_description = "QC, winds in geographic, tilt corrected coordinates, CSAT3s shadow corrected" ; :project_config = "/net/isf/isff/projects/CHATS/ISFF/config/high.xml=4725;" ; :wind3d_horiz_coordinates = "geographic" ; :file_length_seconds = 86400 ; :wind3d_horiz_rotation = 1 ; :wind3d_tilt_correction = 1 ; 2 •The calculated 2-m longwave radiation (Eq. A.15) are in an earlier version of the EOLgenerated 5-min netCDF data file that did not include the shadow-correction. The calculation of longwave radiation is seen in the global attributes of the earlier netCDF file which is: // global attributes: :history = "Created: 2015-07-24 03:16:27 +0000\n", "Updated: 2015-08-04 19:08:06 +0000\n", "Add derived soil properties values.\n", "Applied Philip, 1961 correction to soil heat flux (Gsoil).\n", "Updated: 2015-08-04 19:08:44 +0000\n", "Add derived Gsfc.\n", "Updated: 2015-08-04 19:09:26 +0000\n", "Correct Rnet for wind speed\n", "Updated: 2015-08-04 19:10:04 +0000\n", "Add derived Rlw\n", "Updated: 2015-08-04 19:10:49 +0000\n", "Add derived specific humidity, Q.\n", "" ; :NIDAS_version = "7256" ; :calibration_file_path = "/net/isf/isff/projects/CHATS/ISFF/cal_files/QC,version=1561:/net/isf/isff/cal_files,version=550" ; :dataset = "qc_geo_tiltcor" ; :dataset_description = "QC, winds in geographic, tilt corrected coordinates" ; :project_config = "/net/isf/isff/projects/CHATS/ISFF/config/high.xml=4529M;" ; :wind3d_horiz_coordinates = "geographic" ; :file_length_seconds = 86400 ; :wind3d_horiz_rotation = 1 ; :wind3d_tilt_correction = 1 ; :R_isfs_package_version = "0.0-499" ; :R_isfs_project_version = "" ; :R_version = "R version 3.2.1 (2015-06-18)" ; :lambdasoil_fat_corrected = 1 ; :gsoil_philip_corrected = 1 ; •The flux footprint of the observed turbulent fluxes is considered in Sect 2.1.3 of Bonan et al. (2026). Above-canopy fluxes for winds from the north are likely sampling regions outside of the walnut orchard. No data within the 30-min CSV data file have been eliminated based on footprint considerations. •Most of the data-processing steps used to create the 30-min data are described in ISFS (2025), especially “Corrections to Sensible and Latent Heat Flux Measurements” (https://www.eol.ucar.edu/node/1941). Below we specify which corrections were used and where that correction was applied (either within the 5-min data or in the calculation of the 30-min data): % Sonic Anemometer Tilt-Correction = Yes (planar-fit) [5-min data] % % Sonic Anemometer Transducer Shadowing Correction = Yes [5-min data] % % Schotanus Correction for Sensible Heat Flux = Yes [30-min data] % % Oxygen Correction for Krypton Hygrometers = Yes [30-min data] % % WPL Correction for Water Vapor Flux = Yes [30-min data] % % Spatial Separation of Scalar and Wind Sensors = No % % Condition-dependent values of air density, specific heat % and latent heat of vaporization = Yes [30-min data] % % Mean air T/RH vertically interpolated if <= 4 levels % missing = Yes [30-min data] % 3 3. Data Columns of the CSV Data Files The table below describes which variable is within which column of the CSV data file. Each row below has the following information: (1) column number, (2) a short variable name, (3) variable units, (4) number of missing 30-min data (for May 2007), (5) measurement height. (6) long variable name, (7) sensor type or manufacturer. % Columns are: % % 1-6. Year, Month, Day, Hour, Minute, Sec -- in UTC, Time Stamp Corresponds to center of Averaging Time Period % 07. Decimal Day of Year (UTC) % 008. Rsw_in_16m W/m2 16 16.0 m Incoming Shortwave Radiation Kipp and Zonen CM21 % 009. Rsw_out_16m W/m2 6 16.0 m Outgoing Shortwave Radiation Kipp and Zonen CM21 % 010. Rlw_in_16m W/m2 6 16.0 m Incoming Longwave Radiation Kipp and Zonen CG4 % 011. Rlw_out_16m W/m2 144 16.0 m Outgoing Longwave Radiation Kipp and Zonen CG4 % 012. Rsw_in_2m W/m2 5 2.0 m Incoming Shortwave Radiation Eppley PSP % 013. Rsw_out_2m W/m2 5 2.0 m Outgoing Shortwave Radiation Eppley PSP % 014. Rlw_in_2m W/m2 5 2.0 m Incoming Longwave Radiation Eppley PIR % 015. Rlw_out_2m W/m2 5 2.0 m Outgoing Longwave Radiation Eppley PIR % 016. P_1m Pa 6 1.0 m Barometric Pressure Vaisala PTB220-B % 017. Ta_1.5m K 30 1.55 m Air Temperature Aspirated NCAR/Vaisala Hygrothermometers % 018. Ta_3.0m K 30 3.11 m Air Temperature Aspirated NCAR/Vaisala Hygrothermometers % 019. Ta_4.5m K 30 4.54 m Air Temperature Aspirated NCAR/Vaisala Hygrothermometers % 020. Ta_6.0m K 30 6.07 m Air Temperature Aspirated NCAR/Vaisala Hygrothermometers % 021. Ta_7.5m K 30 7.65 m Air Temperature Aspirated NCAR/Vaisala Hygrothermometers % 022. Ta_9.0m K 30 9.04 m Air Temperature Aspirated NCAR/Vaisala Hygrothermometers % 023. Ta_10.0m K 30 10.04 m Air Temperature Aspirated NCAR/Vaisala Hygrothermometers % 024. Ta_11.0m K 30 11.12 m Air Temperature Aspirated NCAR/Vaisala Hygrothermometers % 025. Ta_14.0m K 30 13.57 m Air Temperature Aspirated NCAR/Vaisala Hygrothermometers % 026. Ta_18.0m K 30 17.52 m Air Temperature Aspirated NCAR/Vaisala Hygrothermometers % 027. Ta_23.0m K 30 22.55 m Air Temperature Aspirated NCAR/Vaisala Hygrothermometers % 028. Ta_29.0m K 30 28.51 m Air Temperature Aspirated NCAR/Vaisala Hygrothermometers % 029. RH_1.5m RH 30 1.55 m Relative Humidity Aspirated NCAR/Vaisala Hygrothermometers % 030. RH_3.0m RH 30 3.11 m Relative Humidity Aspirated NCAR/Vaisala Hygrothermometers % 031. RH_4.5m RH 30 4.54 m Relative Humidity Aspirated NCAR/Vaisala Hygrothermometers % 032. RH_6.0m RH 30 6.07 m Relative Humidity Aspirated NCAR/Vaisala Hygrothermometers % 033. RH_7.5m RH 30 7.65 m Relative Humidity Aspirated NCAR/Vaisala Hygrothermometers % 034. RH_9.0m RH 30 9.04 m Relative Humidity Aspirated NCAR/Vaisala Hygrothermometers % 035. RH_10.0m RH 30 10.04 m Relative Humidity Aspirated NCAR/Vaisala Hygrothermometers % 036. RH_11.0m RH 30 11.12 m Relative Humidity Aspirated NCAR/Vaisala Hygrothermometers % 037. RH_14.0m RH 30 13.57 m Relative Humidity Aspirated NCAR/Vaisala Hygrothermometers % 038. RH_18.0m RH 30 17.52 m Relative Humidity Aspirated NCAR/Vaisala Hygrothermometers % 039. RH_23.0m RH 30 22.55 m Relative Humidity Aspirated NCAR/Vaisala Hygrothermometers % 040. RH_29.0m RH 30 28.51 m Relative Humidity Aspirated NCAR/Vaisala Hygrothermometers % 041. sh_1.5m g/kg 31 1.55 m Specific Humidity Aspirated NCAR/Vaisala Hygrothermometers % 042. sh_3.0m g/kg 31 3.11 m Specific Humidity Aspirated NCAR/Vaisala Hygrothermometers % 043. sh_4.5m g/kg 31 4.54 m Specific Humidity Aspirated NCAR/Vaisala Hygrothermometers % 044. sh_6.0m g/kg 31 6.07 m Specific Humidity Aspirated NCAR/Vaisala Hygrothermometers % 045. sh_7.5m g/kg 31 7.65 m Specific Humidity Aspirated NCAR/Vaisala Hygrothermometers % 046. sh_9.0m g/kg 31 9.04 m Specific Humidity Aspirated NCAR/Vaisala Hygrothermometers % 047. sh_10.0m g/kg 31 10.04 m Specific Humidity Aspirated NCAR/Vaisala Hygrothermometers % 048. sh_11.0m g/kg 31 11.12 m Specific Humidity Aspirated NCAR/Vaisala Hygrothermometers % 049. sh_14.0m g/kg 31 13.57 m Specific Humidity Aspirated NCAR/Vaisala Hygrothermometers % 050. sh_18.0m g/kg 31 17.52 m Specific Humidity Aspirated NCAR/Vaisala Hygrothermometers % 051. sh_23.0m g/kg 31 22.55 m Specific Humidity Aspirated NCAR/Vaisala Hygrothermometers % 052. sh_29.0m g/kg 31 28.51 m Specific Humidity Aspirated NCAR/Vaisala Hygrothermometers % 053. ws_1.5m m/s 31 1.54 m Wind Speed Sonic Anemometer CSAT3 % 054. ws_3.0m m/s 38 3.11 m Wind Speed Sonic Anemometer CSAT3 % 055. ws_4.5m m/s 31 4.54 m Wind Speed Sonic Anemometer CSAT3 % 056. ws_6.0m m/s 35 6.07 m Wind Speed Sonic Anemometer CSAT3 % 057. ws_7.5m m/s 32 7.62 m Wind Speed Sonic Anemometer CSAT3 % 058. ws_9.0m m/s 37 9.02 m Wind Speed Sonic Anemometer CSAT3 % 059. ws_10.0m m/s 32 10.03 m Wind Speed Sonic Anemometer CSAT3 % 060. ws_11.0m m/s 35 11.13 m Wind Speed Sonic Anemometer CSAT3 % 061. ws_12.5m m/s 9 12.54 m Wind Speed Sonic Anemometer CSAT3 % 062. ws_14.0m m/s 29 14.09 m Wind Speed Sonic Anemometer CSAT3 % 063. ws_18.0m m/s 25 18.04 m Wind Speed Sonic Anemometer CSAT3 % 064. ws_23.0m m/s 26 23.07 m Wind Speed Sonic Anemometer CSAT3 % 065. ws_29.0m m/s 23 29.03 m Wind Speed Sonic Anemometer CSAT3 % 066. wd_1.5m deg 31 1.54 m Wind Direction Sonic Anemometer CSAT3 % 067. wd_3.0m deg 38 3.11 m Wind Direction Sonic Anemometer CSAT3 % 068. wd_4.5m deg 31 4.54 m Wind Direction Sonic Anemometer CSAT3 % 069. wd_6.0m deg 35 6.07 m Wind Direction Sonic Anemometer CSAT3 % 070. wd_7.5m deg 32 7.62 m Wind Direction Sonic Anemometer CSAT3 % 071. wd_9.0m deg 37 9.02 m Wind Direction Sonic Anemometer CSAT3 % 072. wd_10.0m deg 32 10.03 m Wind Direction Sonic Anemometer CSAT3 % 073. wd_11.0m deg 35 11.13 m Wind Direction Sonic Anemometer CSAT3 % 074. wd_12.5m deg 9 12.54 m Wind Direction Sonic Anemometer CSAT3 % 075. wd_14.0m deg 29 14.09 m Wind Direction Sonic Anemometer CSAT3 % 076. wd_18.0m deg 25 18.04 m Wind Direction Sonic Anemometer CSAT3 % 077. wd_23.0m deg 26 23.07 m Wind Direction Sonic Anemometer CSAT3 % 078. wd_29.0m deg 23 29.03 m Wind Direction Sonic Anemometer CSAT3 % 079. ustar_1.5m m/s 31 1.54 m Friction Velocity Sonic Anemometer CSAT3 % 080. ustar_3.0m m/s 38 3.11 m Friction Velocity Sonic Anemometer CSAT3 4 % 081. ustar_4.5m m/s 31 4.54 m Friction Velocity Sonic Anemometer CSAT3 % 082. ustar_6.0m m/s 35 6.07 m Friction Velocity Sonic Anemometer CSAT3 % 083. ustar_7.5m m/s 32 7.62 m Friction Velocity Sonic Anemometer CSAT3 % 084. ustar_9.0m m/s 37 9.02 m Friction Velocity Sonic Anemometer CSAT3 % 085. ustar_10.0m m/s 32 10.03 m Friction Velocity Sonic Anemometer CSAT3 % 086. ustar_11.0m m/s 35 11.13 m Friction Velocity Sonic Anemometer CSAT3 % 087. ustar_12.5m m/s 9 12.54 m Friction Velocity Sonic Anemometer CSAT3 % 088. ustar_14.0m m/s 29 14.09 m Friction Velocity Sonic Anemometer CSAT3 % 089. ustar_18.0m m/s 25 18.04 m Friction Velocity Sonic Anemometer CSAT3 % 090. ustar_23.0m m/s 27 23.07 m Friction Velocity Sonic Anemometer CSAT3 % 091. ustar_29.0m m/s 23 29.03 m Friction Velocity Sonic Anemometer CSAT3 % 092. Qh_1.5m W/m2 150 1.54 m Sensible Heat Flux Sonic Anemometer CSAT3 % 093. Qh_3.0m W/m2 150 3.11 m Sensible Heat Flux Sonic Anemometer CSAT3 % 094. Qh_4.5m W/m2 121 4.54 m Sensible Heat Flux Sonic Anemometer CSAT3 % 095. Qh_6.0m W/m2 121 6.07 m Sensible Heat Flux Sonic Anemometer CSAT3 % 096. Qh_7.5m W/m2 125 7.62 m Sensible Heat Flux Sonic Anemometer CSAT3 % 097. Qh_9.0m W/m2 125 9.02 m Sensible Heat Flux Sonic Anemometer CSAT3 % 098. Qh_10.0m W/m2 131 10.03 m Sensible Heat Flux Sonic Anemometer CSAT3 % 099. Qh_11.0m W/m2 131 11.13 m Sensible Heat Flux Sonic Anemometer CSAT3 % 100. Qh_12.5m W/m2 131 12.54 m Sensible Heat Flux Sonic Anemometer CSAT3 % 101. Qh_14.0m W/m2 135 14.09 m Sensible Heat Flux Sonic Anemometer CSAT3 % 102. Qh_18.0m W/m2 132 18.04 m Sensible Heat Flux Sonic Anemometer CSAT3 % 103. Qh_23.0m W/m2 132 23.07 m Sensible Heat Flux Sonic Anemometer CSAT3 % 104. Qh_29.0m W/m2 132 29.03 m Sensible Heat Flux Sonic Anemometer CSAT3 % 105. Qe_1.5m W/m2 150 1.54 m Latent Heat Flux Sonic Anemometer CSAT3 + Krypton Hygrometer % 106. Qe_4.5m W/m2 121 4.54 m Latent Heat Flux Sonic Anemometer CSAT3 + Krypton Hygrometer % 107. Qe_7.5m W/m2 125 7.62 m Latent Heat Flux Sonic Anemometer CSAT3 + Krypton Hygrometer % 108. Qe_10.0m W/m2 131 10.03 m Latent Heat Flux Sonic Anemometer CSAT3 + Krypton Hygrometer % 109. Qe_14.0m W/m2 653 14.09 m Latent Heat Flux Sonic Anemometer CSAT3 + Krypton Hygrometer % 110. Qe_23.0m W/m2 132 23.07 m Latent Heat Flux Sonic Anemometer CSAT3 + Krypton Hygrometer % 111. Gsoil_corr W/m2 5 5 cm depth Soil Heat Flux (with Philips 1961 Correction) REBS HFT-3 % 112. Gsoil_raw W/m2 5 5 cm depth Soil Heat Flux (without Philips Correction) REBS HFT-3 % 113. u_u_1.5m (m/s)^2 31 1.54 m u-wind component variance Sonic Anemometer CSAT3 % 114. u_u_3.0m (m/s)^2 38 3.11 m u-wind component variance Sonic Anemometer CSAT3 % 115. u_u_4.5m (m/s)^2 31 4.54 m u-wind component variance Sonic Anemometer CSAT3 % 116. u_u_6.0m (m/s)^2 35 6.07 m u-wind component variance Sonic Anemometer CSAT3 % 117. u_u_7.5m (m/s)^2 32 7.62 m u-wind component variance Sonic Anemometer CSAT3 % 118. u_u_9.0m (m/s)^2 37 9.02 m u-wind component variance Sonic Anemometer CSAT3 % 119. u_u_10.0m (m/s)^2 32 10.03 m u-wind component variance Sonic Anemometer CSAT3 % 120. u_u_11.0m (m/s)^2 35 11.13 m u-wind component variance Sonic Anemometer CSAT3 % 121. u_u_12.5m (m/s)^2 9 12.54 m u-wind component variance Sonic Anemometer CSAT3 % 122. u_u_14.0m (m/s)^2 29 14.09 m u-wind component variance Sonic Anemometer CSAT3 % 123. u_u_18.0m (m/s)^2 25 18.04 m u-wind component variance Sonic Anemometer CSAT3 % 124. u_u_23.0m (m/s)^2 26 23.07 m u-wind component variance Sonic Anemometer CSAT3 % 125. u_u_29.0m (m/s)^2 23 29.03 m u-wind component variance Sonic Anemometer CSAT3 % 126. v_v_1.5m (m/s)^2 31 1.54 m v-wind component variance Sonic Anemometer CSAT3 % 127. v_v_3.0m (m/s)^2 38 3.11 m v-wind component variance Sonic Anemometer CSAT3 % 128. v_v_4.5m (m/s)^2 31 4.54 m v-wind component variance Sonic Anemometer CSAT3 % 129. v_v_6.0m (m/s)^2 35 6.07 m v-wind component variance Sonic Anemometer CSAT3 % 130. v_v_7.5m (m/s)^2 32 7.62 m v-wind component variance Sonic Anemometer CSAT3 % 131. v_v_9.0m (m/s)^2 37 9.02 m v-wind component variance Sonic Anemometer CSAT3 % 132. v_v_10.0m (m/s)^2 32 10.03 m v-wind component variance Sonic Anemometer CSAT3 % 133. v_v_11.0m (m/s)^2 35 11.13 m v-wind component variance Sonic Anemometer CSAT3 % 134. v_v_12.5m (m/s)^2 9 12.54 m v-wind component variance Sonic Anemometer CSAT3 % 135. v_v_14.0m (m/s)^2 29 14.09 m v-wind component variance Sonic Anemometer CSAT3 % 136. v_v_18.0m (m/s)^2 25 18.04 m v-wind component variance Sonic Anemometer CSAT3 % 137. v_v_23.0m (m/s)^2 26 23.07 m v-wind component variance Sonic Anemometer CSAT3 % 138. v_v_29.0m (m/s)^2 23 29.03 m v-wind component variance Sonic Anemometer CSAT3 % 139. w_w_1.5m (m/s)^2 31 1.54 m w-wind component variance Sonic Anemometer CSAT3 % 140. w_w_3.0m (m/s)^2 38 3.11 m w-wind component variance Sonic Anemometer CSAT3 % 141. w_w_4.5m (m/s)^2 31 4.54 m w-wind component variance Sonic Anemometer CSAT3 % 142. w_w_6.0m (m/s)^2 35 6.07 m w-wind component variance Sonic Anemometer CSAT3 % 143. w_w_7.5m (m/s)^2 32 7.62 m w-wind component variance Sonic Anemometer CSAT3 % 144. w_w_9.0m (m/s)^2 37 9.02 m w-wind component variance Sonic Anemometer CSAT3 % 145. w_w_10.0m (m/s)^2 32 10.03 m w-wind component variance Sonic Anemometer CSAT3 % 146. w_w_11.0m (m/s)^2 35 11.13 m w-wind component variance Sonic Anemometer CSAT3 % 147. w_w_12.5m (m/s)^2 9 12.54 m w-wind component variance Sonic Anemometer CSAT3 % 148. w_w_14.0m (m/s)^2 29 14.09 m w-wind component variance Sonic Anemometer CSAT3 % 149. w_w_18.0m (m/s)^2 25 18.04 m w-wind component variance Sonic Anemometer CSAT3 % 150. w_w_23.0m (m/s)^2 27 23.07 m w-wind component variance Sonic Anemometer CSAT3 % 151. w_w_29.0m (m/s)^2 23 29.03 m w-wind component variance Sonic Anemometer CSAT3 % -------- 4. Appendix A of Bonan et al. (2026) Bonan et al. (2026) contains additional details about the CHATS 30-min data processing. For completeness, Appendix A from Bonan et al. (2026) is listed below; however we strongly urge readers to carefully read Bonan et al. (2026) for the full dataset description. 5 A Appendix: Measurement post-processing A.1 Flux calculation Turbulent fluxes of sensible heat (H) and latent heat (λE) are calculated from the covariance between vertical wind (w) and air temperature (T) or water vapor density (ρv) fluctuations following: H=ρ cpw′T′(A.1) λE =λ w′ρ′ v(A.2) where ρis the mean air density (kg m−3), cpis the mean specific heat of moist air at constant pressure (J kg−1K−1), and λis the mean latent heat of vaporization of water (J kg−1). Note that ρ=ρa+ρvand cp= ( ρacpa+ρvcpv)/ρ, where the subscripts aand vrefer to dry air and water vapor, respectively. Unless otherwise noted, means and fluctuations are calculated over a 30-min period and overbars are only shown on the covariance terms. A.1.1 Covariances As their standard product, NSF NCAR EOL provides 5-min means and covariances (Horst, 2020). The 5-min statistics are used to calculate the 30-min turbulent fluxes where the lowfrequency information needs to be taken into account (ISFS, 2025). For example, to calculate the 30-min covariance between the sonic anemometer temperature (Ts) and vertical wind (w), the following is used: w′T′ s=1 6 6 X j=1 (w′T′ s j+wjTs j)−1 6 6 X j=1 wj1 6 6 X j=1 Ts j(A.3) where w′T′ sis the 30-min covariance, and all the terms on the right-side of Eq. A.3 are 5-min statistics summed over six 5-min periods. Similar expressions are used for w′v′,w′ρ′ v, etc. The effect of using Eq. A.3 compared to taking a simple average is to increase Hby around about 4-5% (the impact is primarily during the daytime when larger-scale eddies are contributing to the flux, which is not captured by a 5-minute averaging window). A.1.2 Additional flux terms and corrections Additional terms are required to the calculate (H) and (λE) from the 30-min covariances. Because temperature and humidity both influence the speed of sound, temperature fluctuations measured by sonic anemometry such as the CSAT3 deployed during CHATS (Campbell Scientific Inc., 2017) are more similar to virtual temperature fluctuations than true air temperature fluctuations. Therefore, we use the Schotanus correction (Schotanus et al., 1983; Foken et al., 2012) to convert sonic temperature into air temperature (i.e., to remove water vapor’s influence). Sensible heat flux is therefore calculated from sonic anemometer temperature (Ts) using: H=ρ cpw′T′ s−0.51 T w′q′(A.4) 6 where qis specific humidity measured with a nearby fast-response humidity sensor (Campbell Scientific Inc., 2021). At CHATS, there were were 13 levels of sonic anemometry on the tower and only 6 levels of krypton hygrometers. Therefore, w′q′comes from the krypton nearest to the sonic anemometer to calculate the Schotanus term. The krypton at the 14-m level was missing a lot of data; therefore, the 23-m krypton is used for the Schotanus corrections of any sonic anemometers at 14 m and higher. On average, the Schotanus correction decreases H by around 5%. Krypton hygrometers use a krypton lamp and Lambert-Beer’s law of absorption to measure water vapor fluctuations in the atmosphere (van Dijk et al., 2003; Foken and Falke, 2012; Campbell Scientific Inc., 2021). The instrument operates at two spectral bands (wavelengths of 116.49 nm and 123.58 nm), and the 116.49 nm wavelength band is attenuated by both water vapor and oxygen. Therefore, a correction for the absorption due to the oxygen is required. For the oxygen-correction, we follow Oncley et al. (2007): w′ρ′ v=w′(ρ′ v)kr +Cko ρa Tw′T′(A.5) where (ρ′ v)kr is the water vapor density output from the krypton hygrometer and the second term is the oxygen correction. Cko is the correction factor for the oxygen absorption, which is calculated via: Cko = (Ko/Kw)(Co Mo Ma ) = 0.23(Ko/Kw)(A.6) where Co= 0.21 is the fraction of oxygen in the atmosphere, and the molecular weights of oxygen and dry air are Mo= 32 and Ma= 28.97 g mol-1, respectively. Koand Kware the krypton hygrometer extinction coefficients for oxygen and water vapor. Kodepends on the path length of the krypton hygrometer, and we use Ko=−0.0034 m3g−1cm−1as recommended in the “Krypton Hygrometers” section of the CHATS data report (CHATS project webpage, 2007). The report also provides the Kwvalue for each individual krypton used in the data processing. For CHATS, the Kwvalues range between −0.143 and −0.152 m3g−1cm−1. The oxygen-correction for λE is small (less than 0.5%). Latent heat fluxes (λE) include the Webb-Pearman-Leuning (WPL) term (Webb et al., 1980; Fuehrer and Friehe, 2002) which is: λE =λ1 + µρv ρahw′ρ′ v+ρv Tw′T′i(A.7) where µ=Ma/Mv= 28.97/18.02 = 1.608 is the ratio of the molecular weight of dry air (Ma) to that of water (Mv), w′ρ′ vis from Eq. (A.5), and w′T′is from the two terms within the brackets in Eq. (A.4). Including the WPL term increases λE by around 1%. A.2 Mean air properties The moist air density (ρ), specific heat of moist air at constant pressure (cp), and latent heat of vaporization (λ) vary with air temperature, pressure, and humidity. These air properties are typically evaluated at the mean air temperature and pressure at the sensor locations; λ, 7 however, should be evaluated at the surface temperature where evapotranspiration occurs. The moist air density is: ρ=ρa+ρv=p−e RdT+e RvT(A.8) where pis pressure, eis the partial pressure of water vapor, Rdis the gas constant for dry air (Rd= 287 J kg−1K−1), and Rvis the gas constant for water vapor (Rv= 461.5J kg−1K−1). In this equation, eis obtained from relative humidity and the saturated vapor pressure (es). For our study, the temperature-dependence of esis calculated using the Goff-Gratch formulation (Goff and Gratch, 1946). The specific heat of moist air at constant pressure is: cp=ρa ρcpa+ρv ρcpv(A.9) which can be equivalently written: cp=cpa1 + ρv ρcpv −cpa cpa (A.10) In this equation, the specific heat of dry air (cpa) has a weak temperature dependence (Garratt, 1992): cpa= 1005 + (T−250)2 3364 ,(A.11) where the value of 1005 J kg-1 K-1 is the specific heat of dry air at T= 250 K. We use cpv= 1850 J kg-1 K-1 for the specific heat of water vapor. Because (cpv −cpa)/cpa ≈0.84,cpsimplifies to: cp≈cpa1+0.84 ρv ρ(A.12) The temperature-dependence of the latent heat of vaporization (Bolton, 1980; Foken, 2017) is: λ= (2.501 −0.00237 T)×106(A.13) where Tis in ◦C. This empirical relationship agrees quite well with WMO-tabulated values of λfrom Letestu (1966). To estimate λat the evaporating surfaces, Twas calculated from the median of the four lowest T/RH sensors on the vertical CHATS tower (all are within the subcanopy). A.3 Air heat storage The heat stored in the air space between the ground and the 23-m measurement height (S) is calculated with (Turnipseed et al., 2002; Leuning et al., 2012): S=ρ cpZ23 m 0 dT dt dz +ρ λ Z23 m 0 dq dt dz (A.14) where air temperature (T) and specific humidity (q) are from the vertical profile of T/RH sensors on the tower, and air density (ρ), specific heat (cp), and latent heat of vaporization (λ) calculations are described above in Sect. A.2. In Eq. (A.14), the first term is the air sensible heat storage and the second term is the latent heat storage. 8 A.4 Radiation measurements Above-canopy radiation was measured at 16 m with a Kipp and Zonen 4-component radiometer (CM21 pyranometers and CG4 pyrgeometers). At 2 m in the subcanopy, EOL-modified (e.g., Delany and Semmer, 1998; Burns et al., 2003) pairs of Eppley model PSP (for shortwave) and model PIR (for longwave) sensors were deployed. To calculate longwave irradiance (RLW) from an Eppley PIR, we follow ISFS (2025) by using: RLW =Rpile +σT4 c−Bσ(T4 d−T4 c) + fRSW (A.15) where Rpile is the thermopile output, σis the Stefan-Boltzmann constant (5.67×10−8 W m−2K−4), Tcis the case temperature, Tdis the dome temperature, Bis the ratio of the dome emissivity to the transmissivity (sometimes called the “dome factor”), fis a correction factor, and RSW is shortwave radiation. Though analytic expressions exist for B, they are usually determined by lab calibration. As described in the ISFS (2025), Band fare both sensor-specific values determined by shading tests in clear-sky conditions. Typical values for Bare from 1.8– 4.0 while for fthey are from 0.5–1.9%. For CHATS, B= 4 was used and the fRSW term was neglected. Additional discussion of Eq. A.15 can be found in Burns et al. (2003) and in the ISFS (2025). 9