Code and Data of "Temporal evolution reveals distinct phytoplankton response dynamics to marine heatwaves"

This workflow is used to detect global marine heatwave (MHW) events, match the events with chlorophyll-a and mixed-layer-depth (MLD) anomalies, normalize event trajectories, and generate the main figures and supplementary analyses in the manuscript.

## Repository contents

### MATLAB scripts

- `A_detect_MHWs.m`: detects MHW events from daily SST and saves the results by longitude blocks as `MHW_*.mat`.

- `B_progress_calculate.m`: calculates SST, MLD, and chlorophyll-a anomalies during each MHW event, normalizes event trajectories, merges block files into `Progress_mhw_event_All.mat`, and calculates grid-scale chlorophyll responses.

- `C_Process_analysis.m`: analyzes event-mean chlorophyll responses and process curves for representative ocean regions.

- `D_peak_time.m`: analyzes the timing of SST, MLD, and chlorophyll anomaly peaks/troughs within normalized MHW event trajectories.

- `E_correlation_analysis.m`: calculates event-scale and grid-scale Spearman correlations among SST, MLD, and chlorophyll trajectories.

- `F_classification.m`: classifies events according to MLD and chlorophyll response types and analyzes the seasonal and interannual variability of different classes.

- `G_Argo_mld.m`: uses Argo/BGC-Argo profiles to support the analysis of temperature, density, chlorophyll, nitrate, and buoyancy frequency.

- `H_FigureSupport.m`: generates supplementary figures and checks the representativeness of the event subset used in the process analysis.

- `Mann_Kendall.m`: Mann-Kendall trend test function.

- `my_yyaxis.m`: helper function for plotting process figures with multiple y-axes.

### Derived data files

This repository only uploads the key intermediate-variable data needed to reproduce the analysis workflow. Larger data files, such as global SST, chlorophyll, MLD data, and global MHW diagnostic results, are not uploaded. Please contact the authors if these files are needed.

- `Progress_mhw_event_All.mat`: merged event-scale dataset and the main input for most analysis scripts.

- `mhw_anomaly_map.mat`, `chl_anomaly_map.mat`, `peak_map.mat`, `R_event.mat`, `curve_grid.mat`, `mhw_character_all.mat`, `mhw_np_count_date.mat`: intermediate analysis results used to reproduce figures.

- `float_match_event.mat`: Argo profile metadata and event labels after matching with MHW events, used by `G_Argo_mld.m`.

- `nino34.long.anom.csv`: climate index input used for ENSO-related analyses.

Some scripts depend on local raw-data paths. These raw data are not included in this upload. Examples include:

- `E:\Data\SST\*.mat`

- `E:\Data\Chl_a\Global_daily_0.25\Chl*.mat`

- `E:\Data\MLD\mld*.mat`

- Argo/BGC-Argo profile data read through local helper functions in the commented code of `G_Argo_mld.m`

The uploaded data are sufficient to reproduce the analysis workflow. To reproduce the complete workflow from the beginning, download the corresponding data from the original data providers and convert them to the variable names and formats required by the scripts. The local paths in the scripts need to be modified before running. Please refer to the data and methods section of the manuscript for details on the data used.

Many figure scripts directly read pre-saved intermediate `.mat` files to reduce computation time. To fully reproduce the workflow from raw inputs, uncomment the corresponding calculation blocks.

## Key variables in `Progress_mhw_event_All.mat`

- `SST_Progress_all`: SST anomaly trajectories for MHW events. Columns 16-65 are the normalized MHW event period; columns 1-15 and 66-80 are the 15-day periods before and after the event, respectively.

- `MLD_Progress_all`: MLD anomaly trajectories, with the same time-column convention as above.

- `Chl_Progress_all`: chlorophyll-a anomaly trajectories, with the same time-column convention as above.

- `mhw_start_all`, `mhw_end_all`: event start and end dates.

- `mhw_lon_all`, `mhw_lat_all`: longitude and latitude of the event grid cells.

- `mhw_dur_all`: event duration.

- `mhw_int_mean_all`: event mean intensity.

## Software dependencies

The scripts are written in MATLAB and use the following toolboxes or external functions:

- MATLAB Statistics and Machine Learning Toolbox, for example `fitlm`, `tcdf`, and related statistical functions.

- `m_map` toolbox, for example `m_proj`, `m_pcolor`, `m_gshhs`, `m_grid`, and `m_scatter`.

- `nclCM` colormap function.

- MHW detection function `detect`.

- Gibbs SeaWater (GSW) Oceanographic Toolbox, used for `gsw_*` functions in the Argo profile calculations.