Research highlights – October 2015
This month in Research Highlights, among others: the latest on flood detection and monitoring, climate drivers of rainfall, assessments of contrarian climate change research and how to integrate satellite data into water observations.
Contrarian climate science papers often mistaken
What is the basis for the conclusions drawn by a small portion of contrarian papers that contradict with the mainstream view on climate change?
In a study titled “Learning from mistakes in climate research”, Dr Rasmus Benested and a team of Australian and European colleagues examine, replicate and test a number of high-profile contrarian papers. In their analyses, errors and mistakes were found in all these contrarian papers, and their type and nature were sorted into different categories.
“One common reason for a flawed conclusion is the failure to include relevant information or context,” says Dr Benested.
The authors highlight the need for independent review of scientific work. “Our paper also makes the case for the importance of independent replication of scientific results.”
Benestad, R., Nuccitelli, D., Lewandowsky, S., Hayhoe, K., Hygen, HO., van Dorland, R. and Cook, J. 2015. Learning from mistakes in climate research Theoretical and Applied Climatology, doi:10.1007/s00704-015-1597-5.
El Nino makes heat waves worse
In parts of Australia that are influenced by El Nino, the hot season starts earlier and has more and hotter heat waves than in areas where its influence is less. This is the result of a new study by Sarah Perkins and Daniel Argüeso of UNSW and Chris White of the University of Tasmania. Overall, the probability of extreme heatwave seasons was found to increase two to four times where El Nino is active.
“This work may allow for better preparedness and planning by those impacted by heatwaves,” explains Dr Perkins. “For example, dry autumn or winter conditions and a developing El Nino will likely put some systems such as bushfire management and public health systems on higher alert in northern and eastern Australia than would average winter conditions or a developing La Nina.”
The authors point out that different physical mechanisms are likely yo affect different heatwave characteristics. Heatwave intensity, duration and the commencement of the heatwave season are less well understood than heatwave frequency.
Dr Perkins notes that the work will also be of value to climate modelling exercises: “models may now be evaluated on their simulated correlations between underpinning mechanisms and Australian heatwaves.”
Perkins, S., Argüeso, D. and White C. J.,. 2015. Relationships between climate variability, soil moisture, and Australian heatwaves, Journal of Geophysical Research, Volume 120, Issue 16, Pages 8144–8164. DOI: 10.1002/2015JD023592
A new flood risk index
New scientific techniques are needed to monitor flood events; understanding their risks and studying their properties including severity, peak flood danger and return periods.
A recent paper proposes a novel index to quantify flood risk. Dr Ravinesh Deo from University of Southern Queensland and a team of Canadian and Korean scientists applied the concept of daily Effective Precipitation.
The method was used to quantify historical flood events in the Brisbane and Lockyer Valley regions. Results showed good skill for monitoring the daily progression of floods, something that was confirmed by records of Brisbane river height, volumetric discharge and dam water levels for key flood events registered in the study region.
“This pilot study demonstrates the practicality of the daily index for flood risk assessments where severity, peak danger, duration or return periods need to be considered,” according to Dr Deo.
Deo Ravinesh C, Byun H-R, Adamowski JF & Kim D-W. (2015). A real-time flood monitoring index based on daily effective precipitation and its application to Brisbane and Lockyer Valley flood event. Water Resources Management. 29(11), 4075-4093. doi:10.1007/s11269-015-1046-3
Natural baseflow indicators can also be used in urban streams
Stream health of urban catchments is declining, and addressing this issue requires a deep understanding of the hydrology of natural and urban catchments. However, so far, it was unclear how hydrological differences between the two types of catchments can be quantified.
In their recent study of urban streamflow dynamics, Dr Perrine Hamel and her colleagues of Monash and Melbourne University found that a common set of indicators can characterize the urban and natural flow regimes.
The research was aimed at guiding practitioners and researchers in the selection of low flow indicators for stormwater management studies.
No studied indicator alone could capture all characteristics of low flows, such as magnitude and timing. “Stormwater management strategies should therefore consider a suite of indicators rather than a single metric,” says Dr Hamel.
Hamel, P., Daly, E. and Fletcher T. (2015), ‘Which baseflow metrics should be used in assessing flow regimes of urban streams?’, Hydrological Processes, Volume 29, Issue 20, pages 4367–4378. DOI: 10.1002/hyp.10475
Satellite flood mapping can improve hydrodynamic modelling
Information on the duration and extent of flood events is important for water resource and environmental management as well as disaster monitoring. Modelling of surface water movement under potential future climate projections also is valuable for determining its environmental impacts.
In a new paper, Dr Cate Ticehurst and her colleagues from CSIRO Land and Water find that flood estimates from satellite observations and modelling can be combined to better understand flooding.
The authors examine potential improvements of the daily MODIS Open Water Likelihood algorithm for flood mapping. They compared the MODIS water fraction maps (500 m pixel size) with Landsat water maps (30 m pixel size) as well as those derived from two-dimensional HD modelling (150 m pixel size).
“Hydrodynamic modelling tools enable detailed analysis of a flood’s progression to a high accuracy however the algorithms are too resource intensive to apply to large catchments,” explains Dr Ticehurst.
“Frequently acquired satellite data can help extend the monitoring and management of flood events at a more regional scale. Conversely, the uncertainty associated with satellite flood mapping can be reduced using the hydrodynamic model.”
Ticehurst, C., Dutta, D., Karim, F., Petheram, C. and Guerschman, J. (2015), ‘Improving the accuracy of daily MODIS OWL flood inundation mapping using hydrodynamic modelling’, Natural Hazards, Volume 78, Issue 2, pp 803-820. DOI: 10.1007/s11069-015-1743-5
Biggest three droughts in NSW attributed to their climate drivers
A new study may have conclusively discovered what caused the the three biggest droughts in southeastern Australia.
Rainfall in eastern Australia are known to be under the influence of four major climate drivers: the El Niño Southern Oscillation, the Interdecadal Pacific Oscillation, the Southern Annular Mode and the Indian Ocean Dipole.
These drivers interact, making it difficult to understand their individual influence on weather. However, a new study for the first time has been able to separately show the influence of each of the four climate drivers, and study their interaction on annual and seasonal rainfall across NSW.
The research team, led by Dr Hiep Nguyen Duc of the NSW Government Climate and Atmospheric Research program, took a different approach from previous studies. “We used Bayesian statistical methods rather than the frequentist regression methods usually applied,” explains Dr Duc. “Our method considers all four climate drivers and their interaction at the same time in a multivariate analysis.”
“Our results provide the resolution to the recent discussion and debate on the cause of three big droughts in south eastern Australia.”
SDuc, H.N., Rivett, K., MacSween, K. and Le-Anh, L. (2015), ‘Association of climate drivers with rainfall in New South Wales, Australia, using Bayesian Model Averaging’, Theoretical and Applied Climatology, pp 1-17. DOI: 10.1007/s00704-015-1622-8
Regional ACCESS forecast model gets Darwin rainfall right, but for the wrong reasons
Regional rainfall forecasts from the Australian ACCESS model over Darwin are good when total rainfall is considered, but only because overestimated rain frequency compensates for underestimated rainfall intensity.
That is the conclusion of a new study by Dr Hahn Nguyen and colleagues of the Centre for Australian Weather and Climate Research.
The team evaluated the 12km horizontal resolution regional ACCESS forecast model against observations from rainfall radar stations (or CPOL). “We characterised rainfall properties by regimes, intensity and frequency of occurrence in order to better understand the model errors,” says Dr Nguyen.
The team compared rainfall during the wet season over Darwin and found that the compensating errors occur throughout the diurnal cycle. There was also a marked land-sea contrast, but it tended to disappear when convection was permitted in the model.
“The errors are mainly due to the convective parameterization set-up that tends to trigger convection too early when solar radiation is highest, which does not match the observations,” explains Dr Nguyen. “One possibility to reduce errors is to increase the horizontal resolution to match the native resolution of the CPOL radar observations.”
Nguyen, H., Protat, A., Kumar, S., Whimpey, M. and Rikus, L. (2015), ‘A regional forecast model evaluation of statistical rainfall properties using the CPOL radar observations in different precipitation regimes over Darwin, Australia’, Quarterly Journal of the Royal Meteorological Society, Volume 141, Issue 691, Pages 2337–2349. DOI: 10.1002/qj.2525
Strength of ENSO influence on rainfall varies over longer time scales
A new analysis of climate model results shows that the strength of the connection between ENSO and eastern Australian rainfall varies widely over time.
Dr Josephine Brown of the Bureau of Meteorology and colleagues of the University of Melbourne analysed six climate models to investigate changes in the connection between El Niño-Southern Oscillation (ENSO) and Australian rainfall over the last 1000 years.
“Summer eastern Australian rainfall is correlated with ENSO in the models, but the strength of the correlation varies from one decade to another, as in the historical record” says Dr Brown.
The changing relationship is due to natural climate processes, and includes decades when the connection breaks down entirely. “Such variability may pose challenges for predicting eastern Australian rainfall using ENSO indices,” concludes Dr Brown.
Brown, J. R., P. Hope, J. Gergis and B. J. Henley (2015), ENSO teleconnections with Australian rainfall in coupled model simulations of the Last Millennium, Climate Dynamics, published online DOI:10.1007/s00382-015-2824-6.