https://doi.org/10.5061/dryad.ht76hdrxc

Citation:
Pérez–Rosales G., Pei, Y.–D., Bennett–Williams J., King T.B., Rummel M., Chung T.H., Wyatt A.S.J. (2026) In situ temperature, stratification, and coral bleaching survey data from the 2022 and 2024 marine heatwaves in Hong Kong [Dataset]. Dryad.
https://doi.org/10.5061/dryad.ht76hdrxc

Data and codes for paper: Pérez–Rosales G., Pei, Y.–D., Bennett–Williams J., King T.B., Rummel M., Chung T.H., Wyatt A.S.J. (2026) Shallow seasonal stratification ameliorates coral bleaching during record‑breaking marine heatwaves in a marginal subtropical system. Coral Reefs.
https://doi.org/10.1007/s00338-026-02835-w

Data archive

• File: Gonzalo_et_al_2026_Data.zip

Compressed archive containing all datasets described in this README,
including the bleaching survey data and environmental monitoring data
from CTD casts, DEFI loggers, fixed loggers, and satellite SST products.

Data/Bleaching Folder:

File: Bleaching_Survey_05022025.csv

• Column Site – study site (Sharp Island, Bluff Island, Tung Ping Chau)
• Column Date – date of survey
• Column Depth – categorical across three depth ranges (1_2m, 2_3m, 4_5m)
• Column Transect – categorical three transects (T1, T2, T3)
• Column Species – categorical species ID taxa, please note that later they were transformed into Genus levels
• Column Size – categorical three sizes (Large, Medium, Small). These were later dropped and not considered in the analysis
• Column Status – multicategory bleaching status (Pigmented, Partial Bleached, Bleached, Dead)
• Column Number Observations – Counting of colonies across each species, size and status

Raw data from the bleaching surveys during August 2024, at the peak of the bleaching, and November 2024 post bleaching.
Please note that the transect surveys were conducted at multi–species–genus levels initially, but that later analysis and results were transformed at genus level only.

File: data_bm_2022.csv

• Column Site – three study sites (Sharp Island, Bluff Island, Tung Ping Chau)
• Column Depth – categorical across two depths (2 and 3)
• Column Species – categorical species ID taxa, directly at genus levels
• Column Status – Binomial (1 = bleached, and 0 = pigmented)
• Column Number Observations – Counting of colonies across each species, size and status
• Column Year – Categorical (2022)

Bleaching data from 2022 (Chung et al 2024 Coral Reefs https://doi.org/10.1007/S00338–024–02533–5). Processed with "Data_Viz_Chung_Data_2022.R" from original data files inside Bleaching/Chung_data/.

File: data_bm_2024.csv

• Column Site – study site (Sharp Island, Bluff Island, Tung Ping Chau)
• Column Depth – categorical across three depth ranges (1, 2 and 3); equivalent to 1_2m, 2_3m and 4_5m
• Column Transect – categorical three transects (T1, T2, T3)
• Column Species – categorical species ID taxa, please note that later they were transformed into Genus levels
• Column Status – Binomial (1 = bleached, and 0 = pigmented)
• Column Number Observations – Counting of colonies across each species, size and status
• Column Year – Categorical (2024)

Bleaching data from 2024. Created from the original Bleaching_Survey_05022025.csv and the Script "Data_Viz_Preliminary.R"

Data/Environment Folder:

Environmental monitoring data from three study sites in Hong Kong:

S1 – Tung Ping Chau  
S18 – Bluff Island  
S21 – Sharp Island

Data/Environment/CTD_Transects:

Folders per Month/Year Cast "Oct2022", "May2022", "Aug2024", "Nov2024"

Each folder contains the date–site identifier and the 11 cast stations stored in a MATLAB (.mat) file with the following variables:

Lat – Latitude (decimal)
Lon – Longitude (decimal)
Station (Cast number)
Transect (Site name code)
data (see below)
metadata (see below)                                                                                        

And inside data, data.physical., there are the following variables:

pressure (dbar) – pressure measured by the sensor, approximately equivalent to water depth
temperature (°C) – seawater temperature
salinity (PSU) – practical salinity derived from conductivity measurements
conductivity1 (mS cm⁻¹) – electrical conductivity measured by the sensor
ec25 (µS cm⁻¹) – electrical conductivity corrected to 25 °C
density (kg m⁻³) – seawater density calculated from temperature and salinity
sigma_t (kg m⁻³) – density anomaly (σt) referenced to atmospheric pressure
chl_flu (ppb) – chlorophyll fluorescence
chl_a (µg L⁻¹) – chlorophyll a concentration
turbidity (FTU) – turbidity measured in formazin turbidity units
do (%) – dissolved oxygen saturation
do_mg (mg L⁻¹) – dissolved oxygen concentration
do_umol (µmol L⁻¹) – dissolved oxygen concentration
battery (V) – logger battery voltage

Inside metadata there are:

SondeName – name of the sonde instrument model
SondeNo – sonde number (instrument identifier – serial number)
Device – device information stored in the instrument metadata
Measurement – measurement information stored in the instrument metadata

Data/Environment/DEFI_loggers

These are loggers mounted on the bleaching transect poles that monitored depth, temperature, and light at the coral level.

One folder: "082024"

File of "DEFI_Average_Transects_202408.xlsx" contains the post–processed data

Column Site – three study sites (Sharp Island, Bluff Island, Tung Ping Chau)
Column Depth (m) – numerical average depths according to mounted loggers
Column Depth standard deviation (m) – numerical depth standard deviation according to mounted loggers
Column Temperature (°C)  – numerical average temperatures for each depth according to mounted loggers
Column Temperature_sd (°C) – Temperature standard deviation 

The raw data are in the form of DEFI2–L, DEFI–D20 and DEFI–T csv files

For DEFI–D20

• Column TimeStamp – Time in format YYYY/mm/dd hh:mm:ss
• Column Pressure (MPa)
• Column Depth (m)
• Column Batt – battery (Volts)

For DEFI–T

• Column TimeStamp – Time in format YYYY/mm/dd hh:mm:ss
• Column Temp – Temperature (°C)
• Column Batt – battery (Volts)

For DEFI2–L

• Column TimeStamp – Time in format YYYY/mm/dd hh:mm:ss
• Column PAR – Photosynthetically Active Radiation (µmol m⁻² s⁻¹)
• Column Batt – battery (Volts)

Data/Environment/Fixed_Loggers

• miniDOT

These data contain temperatures and oxygen from fixed loggers (PME) at different sites and depths

They contain data from 20240428 to post bleaching surveys at 20241105

Data is in a MATLAB file format (.mat) and it contains the following variables:

BV (Volts) – battery voltage
DH – MATLAB cell array containing cumulative Degree Heating metrics calculated from temperature anomaly above the bleaching threshold at multiple temporal scales:
	DH{1} (°C minutes)– Degree heating minutes 
	DH{2} (°C hours)– Degree heating hours 
	DH{3} (°C days)– Degree heating days 
	DH{4} (°C weeks)– Degree heating weeks 
	DH{5} (°C months)– Degree heating months  
DHfilt – MATLAB cell array containing cumulative Degree Heating metrics calculated from a filtered temperature time series in which the semi–diurnal signal has been removed (same units as DH)
DHt – timestamps associated with the Degree Heating metrics (DH), stored as MATLAB serial time (datenum).
DO (mg L⁻¹) – dissolved oxygen concentration 
DOsat (%) – dissolved oxygen saturation
DeploymentType – logger deployment type (BTM = bottom–mounted)                                   
Depth (m) – depth of the logger deployment
HH – cumulative hypoxic exposure (hours where dissolved oxygen falls below 2 mg L⁻¹). This metric was calculated during exploratory analyses and was not used in the bleaching models presented in the manuscript.
HHt – timestamps associated with the HH metric (MATLAB serial date number).
Model – logger model
Q – logger data quality flag indicating measurement reliability (unitless).
SampleInt – logger sampling interval stored in MATLAB serial time units (fraction of a day; e.g., 10 min = 10/1440)
Serial – logger serial number
Site –  study site (Sharp Island, Bluff Island, Tung Ping Chau)
Temp (°C)  – seawater temperature
TempFilt (°C) – seawater temperature time series after filtering to remove the semi–diurnal variability
Time – timestamp of measurement, stored as MATLAB serial time (datenum).
TimeIn – timestamp when the logger was deployed, stored as MATLAB serial time (datenum).
TimeOut – timestamp when the logger was recovered, stored as MATLAB serial time (datenum).

miniDOT loggers were not deployed at S1 Tung Ping Chau

• SBE

These data contain temperatures from fixed loggers (Seabirds) at different sites and depths

They contain data from 20201104 to post bleaching surveys at 20241106

Data is in a MATLAB file format (.mat) and it contains the following variables:

DH – MATLAB cell array containing cumulative Degree Heating metrics calculated from temperature anomaly above the bleaching threshold at multiple temporal scales:
	DH{1} (°C minutes)– Degree heating minutes 
	DH{2} (°C hours)– Degree heating hours 
	DH{3} (°C days)– Degree heating days 
	DH{4} (°C weeks)– Degree heating weeks 
	DH{5} (°C months)– Degree heating months  
DHfilt – MATLAB cell array containing cumulative Degree Heating metrics calculated from a filtered temperature time series in which the semi–diurnal signal has been removed (same units as DH)
DHt – timestamps associated with the Degree Heating metrics (DH), stored as MATLAB serial time (datenum).
DeploymentType – logger deployment type (BTM = bottom–mounted)                      
Depth (m) – depth of the logger deployment                 
Model – logger model                     
SampleInt – logger sampling interval stored in MATLAB serial time units (fraction of a day; e.g., 10 min = 10/1440)                  
Serial – logger serial number                                
Site –  study site                                 
Temp (°C)  – seawater temperature                          
TempFilt (°C) – seawater temperature time series after filtering to remove the semi–diurnal variability 	            
Time – timestamp of measurement, stored as MATLAB serial time (datenum).
TimeIn – timestamp when the logger was deployed, stored as MATLAB serial time (datenum).
TimeOut – timestamp when the logger was recovered, stored as MATLAB serial time (datenum).
d24hrTmax (°C) – maximum seawater temperature within a rolling 24–hour window.
d24hrTmin (°C) – minimum seawater temperature within a rolling 24–hour window.
d24hrTrange (°C) – temperature range within a rolling 24–hour window (d24hrTmax − d24hrTmin).
d24hrTvar (°C²) – temperature variance within a rolling 24–hour window.
d24hrt – timestamps associated with the rolling 24–hour temperature statistics, stored as MATLAB serial time (datenum). 
dailyTavg (°C) – mean seawater temperature calculated for each day.
dailyTmax (°C) – maximum seawater temperature recorded within each day.
dailyTmin (°C) – minimum seawater temperature recorded within each day.
dailyTrange (°C) – daily temperature range calculated as dailyTmax − dailyTmin.
dailyTvar (°C²) – variance of seawater temperature within each day.
dailyt – timestamps associated with the daily temperature statistics, stored as MATLAB serial time (datenum). 

Data/Environment/Satellite_SST

Data file is "HongKong_CoralTemp_0p25deg.mat" and contains the following variables:

SST – Sea surface temperature (°C).
SST_DH – MATLAB cell array containing cumulative Degree Heating metrics calculated from sea surface temperature anomalies above the bleaching threshold at different temporal scales:
	SST_DH{1} (°C days)– Degree heating days 
	SST_DH{2} (°C weeks)– Degree heating weeks 
	SST_DH{3} (°C months)– Degree heating months 
SST_DHt – timestamps associated with the SST_DH metrics, stored as MATLAB serial time (datenum).
SSTt – timestamps associated with the SST time series, stored as MATLAB serial time (datenum).
Temp_avg_day (°C) – Daily average satellite temperature                              
Temp_avg_day_std (°C) – Daily average satellite temperature standard deviation                 
Temp_avg_month (°C) – Monthly average satellite temperature                    
Temp_avg_month_std (°C) – Monthly average satellite temperature standard deviation                   
Temp_daily (°C) – Satellite daily temperature                            
Temp_monthly (°C) – Satellite monthly temperature                      
YEAR                       

Sea surface temperature data obtained from NOAA Coral Reef Watch: https://coralreefwatch.noaa.gov/product/5km/index_5km_sst.php

Software

The R and MATLAB scripts are available at: https://github.com/gonzaloprb/HongKong_Bleaching_Oceanography

See description below:

The R_Scripts are:

• File: Data_Viz_Preliminary.R

This is a massive script mainly for Data exploration and Visualisations

This script includes frequentist statistics as well as the Hurdle models

It generates Fig. 4 and Fig. S3 and Fig. S4

It generates the "data_bm_2024.csv" later used in the Bayesian model

For the Bayesian models better to use Scripts: "Read_Temps_Defi_Bayesian_Model.R" and "Bayesianmodel_Bleaching_Events.R"

• File: Data_Viz_Chung_Data_2022.R

Analyses and extract data from Chung TH, et al (2024) Local conditions modulated the effects of marine heatwaves on coral bleaching in subtropical Hong Kong waters. Coral Reefs 43(5):1235–1247. https://doi.org/10.1007/S00338–024–02533–5

It generates the "data_bm_2022.csv" later used in the Bayesian model

• File: Bayesianmodel_Bleaching_Events.R

Contains the Bayesian Models for the 2022 and 2024 Bleaching Events comparison.

It generates Fig 2d and Fig 2e

• File: Read_Temps_Defi_Bayesian_Model.R

Bayesian modelling of bleaching data according to the mounted DEFI loggers

It generates the Bayesian statistical results

It generates Fig 5 and Fig 6

The best models are Bayes0_Depth and Bayes1_Temp. These are analysed separately.

The Matlab scripts are:

• File: CTD_Reading_Plots_CSV_Casts_Gon.m

It generates the Contour plots of the 11 CTD stations

It generates Fig 3

It runs on the Data/Environment/CTD_Transects

• File: Temp_Loggers_Read.m

It analyses the miniDOT and Seabird fixed logger temperatures

It generates Fig 2 b and c

It generates data for Table 1

The data can be accessed at https://doi.org/10.5061/dryad.ht76hdrxc

• File: Plot_GSHHS_coastline_Gon_Fastcode.m

It plots the SST map

It generates Fig 1

The data can be accessed at https://doi.org/10.5061/dryad.ht76hdrxc

• File: Plot_LongTermSST_Gon.m

It analyses the Satellite data from https://coralreefwatch.noaa.gov/product/5km/index_5km_sst.php

It generates Fig 2 a

Final comment:

If you have any requests, do not hesitate to contact:
gonzalo.prb[@]gmail.com
or
wyatt[@]ust.hk