Methods
An interactive workspace for exploring observed climate indices, their seasonal evolution, and the historical context of highlighted years.
The monitor is designed to compare climate signals across regions and datasets. It includes the major tropical Pacific, Indian, and Atlantic indices used to study El Niño–Southern Oscillation (ENSO) and related climate mode interactions Zhao et al. (2024).
Explore¶
| Monitor | Coverage and datasets | What you can do |
|---|---|---|
| Daily monitor | NOAA OISST v2.1, daily | Compare daily SST, daily anomalies, centered 5-day anomalies, and the daily climatology. |
| Monthly monitor | OISSTv2, ERSSTv5/v6, HadISST, COBE-SST2, and ORAS5 | Compare monthly climate indices across datasets, inspect seasonal composites, and highlight ENSO years. |
How to use the charts¶
- Choose a dataset, index, data mode, and chart view using the left controls.
- Select years in the right panel or click a year in a plot legend; selections are synchronized across panels.
- Hover over a curve to inspect its date and value. Use the plot toolbar to export a figure as PNG or SVG.
- Use the Methods page for the exact index definitions and processing conventions behind each view.
Data and implementation¶
Indices are calculated from gridded observations and reanalysis. The browser renders interactive charts from JSON products generated by the Python scripts; the source datasets and their providers are linked below.
For methodological details, index regions, dataset citations, and ENSO event classification, see the sections that follow.
Data sources¶
| Dataset | Monitor use | Provider and dataset information |
|---|---|---|
| NOAA OISST v2 | Daily monitor; monthly OISSTv2.1 product | NOAA/NCEI OISST v2.1. Daily, quarter-degree optimum-interpolation SST; v2.1 begins in September 1981 and is updated daily. See Huang et al. (2021). |
| ERSSTv5 | Monthly monitor | NOAA PSL ERSSTv5 data page. See Huang et al. (2017). |
| ERSSTv6 | Monthly monitor | NOAA PSL ERSSTv6 data page. Product availability and coverage depend on the local SenPy data installation. |
| HadISST | Monthly monitor | Met Office Hadley Centre HadISST. See Rayner et al. (2003). |
| COBE-SST2 | Monthly monitor | NOAA PSL COBE-SST2 data page. See Hirahara, Ishii, and Fukuda (2014). |
| ORAS5 | Monthly monitor | Copernicus Climate Data Store ORAS5. ORAS5 is an ocean and sea-ice reanalysis produced by ECMWF for the Copernicus Climate Change Service. See Zuo et al. (2019). |
Coverage varies by dataset and may change as source products are updated.
Definitions of climate indices¶
Most indices follow conventional definitions or cited studies, including Zhao et al. (2024), who examine El Niño predictability through interactions among climate modes.
Regional indices are calculated from cosine-of-latitude-weighted SST averages over the stated boxes. Longitudes are written in degrees east or west for readability. For dipole indices, the formula states the order of subtraction; changing that order reverses the sign.
| Index | Operational definition |
|---|---|
| Niño3.4 | SST in 5°S–5°N, 170°W–120°W. |
| Niño3 | SST in 5°S–5°N, 150°W–90°W. |
| Niño4 | SST in 5°S–5°N, 160°E–150°W. |
| Niño1+2 | SST in 10°S–0°, 90°W–80°W. |
| NPMM | SST in 10°N–25°N, 160°W–120°W, following Zhao et al. (2024) and Richter et al. (2022). |
| SPMM | SST in 25°S–15°S, 110°W–90°W, following Zhang et al. 2014. |
| WIO (IOD west) | SST in 10°S–10°N, 50°E–70°E. |
| EIO (IOD east) | SST in 10°S–0°, 90°E–110°E. |
| IOD | WIO − EIO. The western and eastern boxes follow the conventional dipole-mode regions (see Saji et al., 1999). |
| Indian Ocean basin (IOB) | SST in 20°S–20°N, 40°E–100°E (Xie et al., 2009). |
| Southern Indian Ocean west (SIODw) | SST in 25°S–10°S, 65°E–85°E. |
| Southern Indian Ocean east (SIODe) | SST in 30°S–5°S, 90°E–120°E. |
| SIOD | SIODw − SIODe (Jo et al., 2022). |
| Tropical North Atlantic (TNA) | SST in 5°N–25°N, 55°W–15°W (Enfield et al., 1999). |
| ATL3 | SST in 3°S–3°N, 20°W–0° (Nnamchi et al., 2015). |
| South Atlantic subtropical dipole (SASD) | SASD_SW − SASD_NE: 45°S–35°S, 60°W–0° minus 30°S–20°S, 40°W–20°E. |
| TropAve SST | SST over 20°S–20°N, all longitudes. |
| GlobeAve SST | SST over 60°S–60°N, all longitudes. This is the global reference-region average used in the configured SenPy index suite; it is not the North Atlantic AMO index. |
| Southern Ocean SST | 60°S–45°S, all longitudes |
In particular, the PMM indices here are regional SST means, not EOF-based indices.
ENSO event classification¶
ENSO event years and strengths are calculated separately for each dataset from its Niño3.4 series. The monitor provides two methods:
- NOAA running climatology: the NOAA-style 30-year climatology, updated every five years, is used to calculate Niño3.4 anomalies.
- Detrended method: a second-order polynomial trend is removed over 1870–2025, followed by a 21-year running monthly climatology.
For both methods, a centered 3-month mean is passed to the configured
enso_noaa_classifier; its peaks_df() output supplies winter year, peak, and event strength. Consequently, a year highlighted on a January–December chart is the classifier label for that winter and is not, by itself, a display of the event's full two-calendar-year evolution. Event lists can differ between datasets and classification methods.
Daily OISST event highlights use the separately generated OISSTv2_classify_enso.json
product, so the daily SST product contains only SST indices and climatologies.
Processing and displayed modes¶
Daily data¶
Daily OISST indices use no-leap calendar and the 1991–2020 daily climatology. The daily monitor provides:
- Raw SST: regional daily SST index.
- Daily anomaly: daily index minus the climatological value for that calendar day.
- 5-day running anomaly: centered 5-day mean of the daily anomaly.
- Daily climatology: the 365-day 1991–2020 climatological cycle.
Leap days are omitted from the no-leap climatology and plot-year curves.
Monthly data¶
Monthly SST indices are computed with regional average. Monthly anomalies subtract each calendar month's 1991–2020 climatological mean. Detrended anomalies remove a second-order polynomial fit over 1870–2025 where the dataset has sufficient coverage. The seasonal-cycle view displays the 12 monthly climatological values for 1991–2020. Raw mode shows the regional SST index.
References¶
- Zhao et al. (2024), Nature 630, 891–898. DOI
- Richter et al. (2022), npj Climate and Atmospheric Science 5, 94. DOI
- Zhang, Clement & Di Nezio (2014), Journal of Climate 27, 769–783. DOI
- Saji et al. (1999), Nature 401, 360–363. DOI
- Xie et al. (2009), Journal of Climate 22, 730–747. DOI
- Jo et al. (2022), Nature Communications 13. DOI
- Enfield et al. (1999), Journal of Geophysical Research: Oceans 104.
- Nnamchi et al. (2015), Nature Communications 6, 8895. DOI
- Huang et al. (2017), Journal of Climate 30, 8179–8205. DOI
- Huang et al. (2021), Journal of Climate 34, 2923–2939. DOI
- Rayner et al. (2003), Journal of Geophysical Research 108(D14), 4407. DOI
- Hirahara, Ishii & Fukuda (2014), Journal of Climate 27, 57–75. DOI
- Zuo et al. (2019), Ocean Science 15, 779–808. DOI