Search Results

You are looking at 1 - 3 of 3 items for :

  • Author or Editor: J. Eischeid x
  • Journal of Climate x
  • Refine by Access: All Content x
Clear All Modify Search
Porathur V. Joseph, Jon K. Eischeid, and Robert J. Pyle


The long-term mean date of the monsoon onset over Kerala (MOK) varies between 30 May and 2 June according to different estimates, with a standard deviation of 8–9 days. The earliest date of MOK, and the most delayed one, during the last 100 years differ by 46 days (7 May and 22 June, respectively). MOK switches on a spatially large and intense convective heat source over south Asia, lasting from June to September, whose moisture supply is made available through the cross-equatorial low-level jet stream.

Superposed epoch analysis of 10 years of outgoing longwave radiation (OLR) data shows that MOK is a significant stage in the evolution of the OLR field in the tropics of the eastern hemisphere. At the time of MOK there is increased convection in a band about 5–10 degrees wide meridionally, extending from the south Arabian Sea to south China, and convection is suppressed all around, particularly in the western Pacific Ocean. In 1983 when MOK was delayed by about 3 pentads, OLR data showed that the boreal spring-to-summer migration of the equatorial convective cloudiness maximum (ECCM), both westward and northward, was also delayed. The delayed MOK is accompanied by delays in the northwestward movement of ECCM and is confirmed by an analysis of long-term data of southwest Pacific tropical cyclones.

Of the 22 years between 1870–1989 when MOK was delayed by 8 days or more, 16 casts were associated with a moderate or strong El Niño. Of the 13 strong El Niños during the same period, 9 were associated with moderate-to-large delays in MOK. Delays preferentially occurred in the year +1 of an El Niño, where year 0 is the growing phase of the El Niño in sea surface temperature (SST).

Analysis of the SST field has shown that delayed MOK is associated with warm SST anomalies at and south of the equator in the Indian and Pacific oceans and cold SST anomalies in the tropical and subtropical oceans to the north during the season prior to the monsoon onset (i.e., March to May). It is hypothesized that such SST anomalies over the Indian and Pacific oceans (generally found associated with El Niño, either in year 0 or year +1 or in both) cause the interannual variability of the MOK through their action in affecting the timing of the northwestward movement of the ECCM.

Full access
M. Hoerling, J. Barsugli, B. Livneh, J. Eischeid, X. Quan, and A. Badger


Upper Colorado River basin streamflow has declined by roughly 20% over the last century of the instrumental period, based on estimates of naturalized flow above Lees Ferry. Here we assess factors causing the decline and evaluate the premise that rising surface temperatures have been mostly responsible. We use an event attribution framework involving parallel sets of global model experiments with and without climate change drivers. We demonstrate that climate change forcing has acted to reduce Upper Colorado River basin streamflow during this period by about 10% (with uncertainty range of 6%–14% reductions). The magnitude of the observed flow decline is found to be inconsistent with natural variability alone, and approximately one-half of the observed flow decline is judged to have resulted from long-term climate change. Each of three different global models used herein indicates that climate change forcing during the last century has acted to increase surface temperature (~+1.2°C) and decrease precipitation (~−3%). Using large ensemble methods, we diagnose the separate effects of temperature and precipitation changes on Upper Colorado River streamflow. Precipitation change is found to be the most consequential factor owing to its amplified impact on flow resulting from precipitation elasticity (percent change in streamflow per percent change in precipitation) of ~2. We confirm that warming has also driven streamflow declines, as inferred from empirical studies, although operating as a secondary factor. Our finding of a modest −2.5% °C−1 temperature sensitivity, on the basis of our best model-derived estimate, indicates that only about one-third of the attributable climate change signal in Colorado River decline resulted from warming, whereas about two-thirds resulted from precipitation decline.

Full access
J. K. Eischeid, M. P. Hoerling, X.-W. Quan, and H. F. Diaz


Hawaii’s recent drought is among the most severe on record. Wet-season (November–April) rainfall deficits during 2010–19 rank second lowest among consecutive 10-yr periods since 1900. Various lines of empirical and model evidence indicate a principal natural atmospheric cause for the low rainfall, mostly unrelated to either internal oceanic variability or external forcing. Empirical analysis reveals that traditional factors have favored wetness rather than drought in recent decades, including a cold phase of the Pacific decadal oscillation in sea surface temperatures (SSTs) and a weakened Aleutian low in atmospheric circulation. But correlations of Hawaiian rainfall with patterns of Pacific sea level pressure and SSTs that explained a majority of its variability during the twentieth century collapsed in the twenty-first century. Atmospheric model simulations indicate a forced decadal signal (2010–19 vs 1981–2000) of Aleutian low weakening, consistent with recent observed North Pacific circulation. However, model ensemble means do not generate reduced Hawaiian rainfall, indicating that neither oceanic boundary forcing nor a weakened Aleutian low caused recent low Hawaiian rainfall. Additional atmospheric model experiments explored the role of anthropogenic forcing. These reveal a strong sensitivity of Hawaiian rainfall to details of long-term SST change patterns. Under an assumption that anthropogenic forcing drives zonally uniform SST warming, Hawaiian rainfall declines, with a range of 3%–9% among three models. Under an assumption that anthropogenic forcing also increases the equatorial Pacific zonal SST gradient, Hawaiian rainfall increases 2%–6%. Large spread among ensemble members indicates that no forced signals are detectable.

Restricted access