Download presentation
Presentation is loading. Please wait.
Published byJoel Quinn Modified over 8 years ago
1
Acknowledgment: This research was carried out at the Jet Propulsion Laboratory, California Institute of Technology, under contract with NASA. ©Copyright 2012 California Institute of Technology. Hartmut Aumann and Alexander Ruzmaikin California Institute of Technology, Jet Propulsion Laboratory, Pasadena, CA 91109 Trends in the Frequency of Severe Storms using nine years of AIRS data Summary and Conclusions There is consensus that global warming will change the frequency and intensity of severe weather events. There is no consensus if global warming will increase the intensity of severe storms, their frequency, or both. We use 9 years of AIRS data to explore this question. Most of the DCC are aggregated in clusters of 6 or more DCC. A DCC cluster is a collection of DCC which are separated by no more than one AIRS footprint (13 km at nadir). Some clusters have over 1000 DCC. Every day AIRS observes a certain class of cold clouds where the brightness temperature in strong waterlines is warmer than the brightness temperature in atmospheric window channels. The example above is from 1:30 AM local time overpass of AIRS on 16 March 2012. There were 3473 DCC. There were 58 clusters with more than 6 members. A 23 DCC cluster was at E131.7, N12.0 and Darwin, Australia, reported severe flooding. For every day over the past 10 years we used AIRS to identify DCC. This gives time series of daily count of DCC. We then ran a cluster analysis for each day (day overpass and night overpass separately). This gives us a time series of a count of clusters (with more than 6 DCC in a cluster). We used this data to evaluate anomaly trends in the count of DCC and the count of clusters on a global and regional scale. Typically we find 6000 DCC globally each day, most in the tropical zone, but some as for north at 45N latitude. Based on AMSRe (on the same S/C as AIRS), DCC are typically associated with a 4 mm/hr rain rate and 45% convective faction. Each DCC represents active deep convection (45%) or remnants of deep convection. DCC are historically associated with severe weather events. There was a 39 DCC cluster at 90W, N36.6 over the state of Illinois, USA. Severe hail and 75 mph wind gusts were reported. A quick google search, e.g. “16 March 2012 storm” for clusters with more than 30 elements, has a 80% hit-rate. The trend in the global DCC count is 0.03 ±0.12 %/year. The trend analysis defines a DCC cluster as a cluster with 6 or more DCC. The trend in the global DCC cluster count is +0.49 ±0.14 %/year. Next we separate DCC and clusters of DCC by land and ocean. The separation of DCC into land and ocean reveals a strong multiannual variability, which is moderately anti-correlated (r=-0.58). In the past ten years the DCC count decreased by 1%/year for the ocean, while it increased by 2%/year for land. In the past ten years the cluster size for ocean has decreased by -0.83±0.15%/year while for land it has increased by 0.47±0.22%/year. For land the observation that the number of DCC has increased, the number of clusters has increased and the mean cluster size has increased. The count of DCC and their mean size has increased. The count of DCC and the mean size of DCC clusters are proxies for the severity of storms. Are the changes observed over the past ten year part of a climate signal, or are they manifestations of decadal weather variability? For this we turn to the nino34 index, also known as MEI and ONI. It is available from the Hadley Center since 1870. When analyzed with a ten year window, the nino34 index has a strong negative trend. The nino34 index and the frequency of DCC for land are highly anti-correlated (r=-0.79). The nino34 index and the cluster size over land are weakly anti- correlated (r=-0.12). While the frequency of DCC has remained essentially fixed, the frequency of clusters has increased. The global mean cluster size of 40 DCC/cluster has decreased by -0.46±0.12 %/year. We overlay the negative of the nino34 (also known as MEI and ONI) on the land DCC anomaly. The correlation is -0.79. The DCC frequency and cluster size are well significantly correlated (r=0.47) with the nino34 index for ocean. The cluster count is essentially not correlated with nino34 (r=0.02). DCC and clusters of DCC can be used as proxies for severe storms. Over the past 9 years the frequency of DCC, the frequency of clusters and the size of clusters have not changed significantly. The frequency of DCC, the frequency of clusters and the size of clusters have increased significantly over land, but decreased over ocean, such that the change nearly cancels globally. The observed changes over ocean are correlated with the nino34 index, while they are anti- correlated for land. We conclude that the trends in the frequency of storms observed during the past 9 years are related to decadal variability. This results in the transfer of moisture and clouds, including severe storms, from one area of the globe to another. This is correlated with the nino34 index, but has little to do with climate change.
Similar presentations
© 2024 SlidePlayer.com. Inc.
All rights reserved.