Other Scholarly Content
 

Zonal and Meridional Discontinuities and Other Issues with the HadISST1.1 Dataset

Public Deposited

Downloadable Content

Download PDF
https://ir.library.oregonstate.edu/concern/defaults/kw52j9632

Descriptions

Attribute NameValues
Creator
Abstract
  • 3) The standard deviations of the SST anomalies in the HadISST1.1 dataset are 0.2–0.4°C smaller prior to 1949 than after 1949. This nonstationarity could complicate interpretations of the long HadISST1.1 data record for some studies of climate variability. It is likely attributable mostly to the improved sampling of SST observations in the later time period.
  • 4) There is a large zonal discontinuity of SST in the HadISST1.1 dataset along 179.5°W that results in SST variations that are typically smaller in magnitude by about 0.2°C compared with the adjacent 1° grid cells to the west and east. It appears that this discontinuity becomes noticeable abruptly in 1949.
  • 1) The HadISST1.1 maps are much smoother than maps of SST from the Reynolds et al. (2007) Daily Optimal Interpolation SST (OISST) dataset. This is attributable at least partly to the coarser grid resolution of 1°×1° for HadISST1.1 compared with 0.25°×0.25° for OISST. The lower resolution of the HadISST1.1 dataset is not a major issue for many studies of large-scale climate variability.
  • 5) There are additional zonal discontinuities of SST at very regular intervals of 4° and 2° of longitude in, respectively, the pre-1949 and post-1949 portions of the HadISST1.1 data record. These discontinuities in zonal first differences of SST are manifest as step-like structures in SST itself, with typical jumps of 0.1°C prior to 1949 and 0.05°C after 1949.
  • While all of the above issues are significant concerns, we show that the HadISST1.1 dataset is useful for studies of the large-scale signals associated with short-term climate variability (interannual to decadal time scales). The dataset is inadequate, however, for studies that require information about spatial derivatives of the SST field. An example is investigation of ocean-atmosphere interaction on oceanic mesoscales of 100–1000 km.
  • A detailed analysis of the Version 1.1 Hadley Centre Sea Ice and Sea Surface Temperature (HadISST1.1) dataset has identified numerous problems, some of which limit the utility of the dataset for certain scientific applications. The issues identified from our analysis are:
  • 2) The standard deviations of the SST anomalies are 0.2–0.4°C smaller in the HadISST1.1 dataset than in the OISST dataset. This is likely also attributable at least partly to the coarser grid resolution of the HadISST1.1 dataset. However, the magnitudes of these differences seem larger than could be accounted for solely by the coarser resolution of the HadISST1.1 dataset.
  • 6) There are also meridional discontinuities at intervals of 4° and 2° of latitude in, respectively, the pre-1949 and post-1949 portions of the HadISST1.1 data record. The amplitudes of the meridionally oscillating striations in the standard deviations of meridional first differences are as large or larger than the amplitudes of the zonally oscillating striations in the standard deviations of zonal first differences.
Resource Type
Date Available
Date Issued
Rights Statement
Funding Statement (additional comments about funding)
  • NASA
Peer Reviewed
Language
Replaces

Relationships

Parents:

This work has no parents.

Items

Thumbnail Title Date Uploaded Visibility Actions
file details Chelton___Risien_Report_on_HadISST1.1_Issues.pdf 2017-07-24 Public Download