Australia: The Land Where Time Began

A biography of the Australian continent 

Pan-Arctic Melt Onset – Recent Changes from Satellite Passive Microwave Measurements

A new passive microwave (PMW) melt onset retrieval algorithm that is based on temporal variations in the differences of the brightness temperature between 19 and 37 Ghz has been shown to be effective as radar (e.g. QuikScat) measurements. Improved estimates of melt that are linked more closely to snow-off dates that have been observed has been demonstrated for the PMW technique than in previous studies. According to Wang et al. combining estimates on land and sea ice for the entire PMW record produced an integrated pan-Arctic (north of 50^o N) melt onset date (MOD) dataset. Significant trends of 2⁓3 days/decade to earlier MOD over the period 1979-2011are concentrated mainly over the Eurasian land sector of the Arctic, which is consistent with changes in the extent of snow cover in spring that has been observed with visible satellite data. Spring surface air temperature largely drives the variability and change in the onset of melt, with significant influence from low frequency modes of atmospheric circulation.

In recent decades there has been rapid warming and increasing precipitation in winter in the Arctic (Trenberth et al., 2007; Min et al., 2008). E.g., there has been increasing depth of snow in many regions of Eurasia (EUR; Bulygina et al., 2009), at the same time there have been widespread decreases in snow depth in North America (NA; Callaghan et al., 2011). Based on the weekly snow chart dataset from NOAA, the extent of the spring snow cover (SCE) has exhibited more negative trends in Eurasian Arctic than the North American Arctic (Brown et al., 2010; Derksen & the Brown, 2012). This is in contrast to trends in spring snowmelt records that are derived from satellite passive microwave (PMW) data that show a more consistent pan-Arctic response (Tedesco et al., 2009). It has been found that surface air temperature (SAT) exerted the most significant influence on the interannual variability in spring SCE (Brown et al., 2010; Derksen & Brown, 2012). The seasonal strength of the Arctic Oscillation Index (OA), however, was found to also influence the interannual variability in the date of melt onset (MOD) on the Arctic sea ice and across the Eurasian terrestrial Arctic (Drobot & Anderson, 2001; Belchansky et al., 2001; Tedesco et al., 2009).

Microwave satellites are effective tools for detecting changes in the dynamics of snowmelt across the Arctic as a result of their high sensitivity to liquid water in snow and generally absorbance of the cover issues faced by visible satellite imagery (Wang et al., 2008, Markus et al., 2009). And integrated pan-Arctic melt onset dataset was developed (Wang et al., 2011) by combining active and passive microwave-derived estimates for the northern high latitude land surface, ice caps, large lakes, and sea ice for the period 2000-2009. The integrated dataset allows the examination of melt dynamics in a full pan Arctic context, as well as the exploration of the interaction between terrestrial and marine components of the cryosphere, which is a major advantage. The dataset is also useful for the evaluation of model simulations in the spring period, a time during which models exhibit significant spread in the simulation of snow cover (Slater et al., 2001). It also provides independent validation of trends in the extent of snow cover observed with visible satellite dataset (e.g. Derksen & Brown, 2012) as it has been shown that satellite-derived MOD estimates correlate significantly with observed snow-off dates (Wang et al., 2008).

The PMW data have a coarser resolution (25 km) and have previously been found to be less sensitive to melt compared to enhanced resolution active microwave  QuikScat (QS) data (⁓5 km). The QS data are available for only between June 1999 and November 2009, because the antenna stopped working. In order to take advantage of a longer melt detection technique the development of a new PMW melt detection technique is necessary. Therefore, the objectives of this study were as follows:

·        To present a new improved melt detection algorithm for PMW data that is capable of identifying multiple melt events and produce estimates of melt that are closely correlated to the end of snow season (as was developed for QS in Wang et al., 2008);

·        To extend the 10-year pan-cryosphere integrated melt dataset of Wang et al. (2011) to the whole satellite PMW data record;

·        To analyse trends in MOD across the Arctic land and sea ice and examine the relative roles of warming and atmospheric circulation in the variability that has been observed and changes in the MOD over the period 1979-2011.

Discussions and conclusions

In this study Wang et al. developed an algorithm that is capable of separation of multiple early melt events from the main melt event by the use of satellite PMW measurements. The new algorithm shows improved estimates of melt that are more closely linked to snow-off dates that have been observed  than a previous pan-Arctic study (Tedesco et al., 2009). The TbD from 10 and 37 GHz PMW measurements proved to be as effective as σ^o from QS for the detection of multiple melt events. The PMW algorithm performs better in the dense forest and high relief areas than data from QS even though the spatial resolution is coarser on account of large range of TbD. In this study the correlation between MED and snow-off is not as strong as that found by Wang et al. (2008). The end date of melting snow cannot be directly identified from TbD is one of the main reasons, while σ^o from QS can detect melt end explicitly (Wang et al., 2008).

In this study the spatial pan-Arctic trends from PMW data are consistent with results that are PMW derived for Arctic sea ice (Maksimovich & Vihma, 2012) and the land surface of the Northern Hemisphere (Kim et al., 2012) with trends that are mostly nonsignificant with the exception of over a narrow band on central Arctic sea ice and land areas of Eurasia. It was suggested (Kim et al., 2012) that the lack of significant trends may be related to the large interannual variability of MOD. It was documented in a previous study (Tedesco et al., 2009) that pan-Arctic trends are towards earlier MOD, though the results of Wang et al. indicate greater spatial variability with the strongest trends being towards earlier MOD of the Eurasian land sector of the Arctic, which is consistent with SCE trends derived from visible satellite data that have been reported (Derksen & Brown, 2012).

It has been found (Tedesco et al., 2009) that up to 50% of the variance in MOD in Eurasia can be explained by the seasonal strength of the AO index during 1979-2008. The study by Wang et al. (2013) shows that for the period 1979-2011, the AO index is correlated significantly only with MOD in the subarctic zone of North America and Eurasia. The PNA index has an influence on the variability in MOD that is relatively strong in the Subarctic of North America. The spring SAT has a much stronger influence of the variability of MOD in both sectors of the Arctic, when compared to that of the AO and the PNA, which is consistent with the findings in other studies for spring SCE (Brown et al., 2010; Derksen & Brown, 2012). MOD trend are largely driven by changes in the spring SAT over the period of 1979-2011, with only a small contribution characterised by the atmospheric circulation indices. Continued monitoring of the variability and change in spring MOD from satellite [data] will provide useful information on the response of the cryosphere to amplified warming in the Arctic, given the close link between MOD and SAT.

Sources & Further reading

1. Wang, L., et al. (2013). "Recent changes in pan-Arctic melt onset from satellite passive microwave measurements." Geophysical Research Letters 40(3): 522-528.