Observations

Can a 3D radar composite reliably represent ZDR columns?

The Social MetworkLeave a comment

Brian Lo – brian.lo@pgr.reading.ac.uk 

Differential reflectivity (ZDR) is the difference in measured backscatter from emitted radio waves in the horizontal and vertical polarisations. It is an observable available from dual-polarisation radars. Conventional single-polarisation radars usually offer reflectivity ZH only which is the intensity of backscatter in the horizontal polarization available from conventional radars. The addition of measuring ZDR allows for hydrometeor type classification. In other words, we could tell between large round tumbling hail which gives near zero ZDR from large oblate raindrops that give highly positive ZDR (Kumjian, 2013a). Strong updrafts contribute to severe convective development by lofting large hydrometeors like raindrops into higher parts of the storm, giving ZDR column signatures. A differential reflectivity (ZDR) column is defined as “a region of enhanced ZDR as situated above the 0◦C level” (Kumjian, 2013b) and are known to be useful in informing forecast warning decisions (e.g. Kuster et al., 2019, 2020).

The UK Met Office has fully upgraded all 15 C-band radars as of January 2018 to have dual-polarisation capabilities. The Met Office also composites data from this radar network to provide three-dimensional gridded products covering the entirety of the UK (Scovell and al-Sakka, 2016). Whereas a single radar would only be able to detect hydrometeors as high as its highest scan elevation, thus leaving the so-called “cone of silence” aloft closest to the radar, the composite permits nearby radars to fill in these regions of missing observations. To harness the greater spatial domain of the 3D radar composite constituting data from multiple overlapping radars, the composite was upgraded to include ZDR to investigate the operational potential of using ZDR columns. But, how do we know ZDR columns can be reliably detected within this 3D radar composite? 

The work described in this post is to verify ZH and ZDR generated by the Met Office compositing process against range-height indicator (RHI) scans from Chilbolton Advanced Meteorological Radar (CAMRa), otherwise known as the world’s largest steerable meteorological research radar (see Figure 1). RHI scans are carried out by varying a radar antenna’s elevation angle but with the azimuth angle held constant. ZDR columns are often narrow features that may challenge the limited resolution (1km in the horizontal) of the radar composite.

Figure 1: The 25 m antenna of the Chilbolton Advanced Meteorological Radar (CAMRa) located at the Chilbolton Observatory. 

CAMRa is a suitable truth owing to it being well-calibrated to within 0.1dB of ZDR, its extremely narrow beamwidth of 0.28◦, high range resolution of 75m and high resolution elevations of 0.11deg within RHIs. In contrast, the composite is made of operational radar data of lower resolution and the compositing process could further degrade the accuracy of the data. Thus its ZDR output has to be verified.

Vertical cross sections of the radar composite

Figures 2a and 2d are RHI scans carried out by CAMRa, covering elevations from 0.02 to 10.0◦. These two figures captured an evolving convective system on 7 June 2016 at 1651Z (Figure 2e) and on 1 October 2019 at 1526Z (Figure 2f). The fine range resolution of 75m captured multiple intense reflectivity cores exceeding 40 dBZ with accompanying overshooting tops.

Figures 2b and 2e are pseudo-RHIs produced from the compositing process of the Chenies and Thurnham operational C-band radars. Both radars were chosen for the compositing process as their overlapping sampling regions offered coverage for the convective system scanned by CAMRa. Instead of generating the usual 3D composite with 1 km and 500 m of horizontal and vertical grid spacing respectively, the compositing software was modified in this verification process to interpolate C-band radar data onto a 2D grid along the CAMRa scan azimuth with the same grid resolutions thus producing the so-called pseudo-RHI plots. 

Figure 2: RHI plots of radar reflectivity ZH scanned by CAMRa (a,d) and corresponding pseudo-RHI plots derived from the original Met Office radar compositing process (b,e) and with azimuthal correction applied (c,f). The top and bottom rows each corresponds to observations on 07 June 2016 but at 1603Z and 1651Z respectively. 

However, the preliminary inspection of the pseudo-RHI plots reveals a serious vertical discontinuity of interpolated data. Such an issue would disrupt the automatic detection of vertically extending signatures such as ZDR columns. This problem is seen at around 70 and 100km down range in Figure 1b and 90km down range in Figure 2e. The displacement observed here suggests that the spatial location of storms could be misrepresented on the order of 5 km. What could have caused this problem?

Correction of spatial location of radar beams in compositing software 

Through scrutiny of the Met Office compositing software, I found an error with how radar azimuths were used in the compositing process. In the pre-exisiting compositing software,  radar azimuths would be formulated as azimuthal equidistant projection coordinates relative to the radar site, then directly transformed onto British National Grid coordinates. However, it was overlooked that Met Office radar azimuths are recorded with respect to British National grid north, whereas azimuthal equidistant projection coordinates require azimuths to be relative to true north. The deviations of up to a few degrees between the norths can lead to a horizontal displacement of radar data of at least a few kilometres at a range of 100km from a radar site. Lesson learnt: Attention to small details can have a large impact later on! 

To fix this problem, grid convergence (Ordnance Survey, 2018) is added to all radar azimuths such that radar azimuths are adjusted with respect to true north before undergoing transformation into British National Grid coordinates. The effect of implementing such a correction can be seen in Figures 1c and 1f, where reflectivity values interpolated from two separate C-band radars result in a vertically continuous intense reflectivity core. There could still be mismatches on a smaller scale owing to radar scans happening at different times while the storm was being advected. With the radar composite corrected, is ZDR well represented? 

Visual comparison of ZDR

Both Chenies and Thurnham radars used for generating pseudo-RHIs were upgraded to have dual polarimetric capability since March 2013. This allows the generation of ZDR pseudo-RHI plots as shown in Figures 2c and 2b, which can be qualitatively compared with CAMRa scans in Figures 2a and 2d on 7 June 2016 at 165136Z and on 1 October 2019 at 152641Z respectively.

Figure 3: RHI plots of radar differential reflectivity ZDR scanned by CAMRa (a,b), corresponding pseudo-RHI plots derived from the radar composite (c,d) and MAXDBZ plan views in (e,f). In the plan views, the red cross marks the position of CAMRa. The black line is the azimuth of CAMRa for the RHI scan. Black dots are separated by 20km with the first and last black dot corresponding to the plotted range of the RHIs. The top and bottom rows each corresponds to observations on 07 June 2016 at 1651Z and 01 Oct 2019 at 1526Z respectively. The freezing height was 3.0 km in the June case and 2.2 km in the October case. 

Considering the observed ZDR column is approximately 95 km down range from the radar, a displacement of 0.6◦ is 1km of distance in the horizontal. Such a distance corresponds to the sampling resolution of the radar composite. The C-band operational radars also have a wider beamwidth of 1.1◦ and are unable to observe fine details unless the ZDR column is situated close to one of the radars. Thus, the discrepancy in intensity and height in Figure 2 is expected, owing to the differences in sampling resolutions between CAMRa and the radar composite. The operational radar composite, which combines measurements from radars at various ranges, is capable of detecting ZDR column features at a coarser resolution, whereas CAMRa is used to study fine details of the column structures with high precision in individual case studies. 

We have shown that outputs from CAMRa captured sub-kilometre features such as the width of ZDR columns and their horizontal structures within a cell are too fine to be resolved by the composite. Despite the resolution limitations of operational radars and having done other tests not mentioned in this post, we are confident that the radar composite can be exploited to reliably capture the presence of ZDR columns at a horizontal spatial resolution of 1 km alongside an indication of their maximum heights. 

References

Kumjian, M. R. (2013a). “Principles and Applications of Dual-Polarization Weather Radar. Part I: Description of the Polarimetric Radar Variables”. Journal of Operational Meteorology 1.20, pp. 243–264. doi:10.15191/nwajom.2013.0120 

Kumjian, M. R. (2013b). “Principles and applications of dual-polarization weather radar. Part II: Warm- and cold-season applications”. Journal of Operational Meteorology 1.20, pp. 243–264. doi:10.15191/nwajom.2013.0120 

Kuster, C. M. et al. (2019). “Rapid-update radar observations of ZDR column depth and its use in the warning decision process”. Weather and Forecasting 34.4, pp. 1173–1188. doi: 10.1175/WAF-D-19-0024.1 

Kuster, C. M. et al. (2020). “Using ZDR Columns in Forecaster Conceptual Models and Warning Decision Making”. Weather and Forecasting, pp. 1–43. doi: 10.1175/WAF- D-20-0083.1 Ordnance Survey (2018). A Guide to Coordinate Systems in Great Britain. Accessed: 16-1-2024 

Scovell, R. and H. al-Sakka (2016). “A Point Cloud Method for Retrieval of High-Resolution 3D Gridded Reflectivity from Weather Radar Networks for Air Traffic Management”. Journal of Atmospheric and Oceanic Technology 33.3, pp. 461–479. doi: https://doi.org/10.1175/JTECH-D-15-0051.1 

Designing a program to improve data access for my PhD project

The Social MetworkLeave a comment

Caleb Miller – c.s.miller@pgr.reading.ac.uk

In my project work, I regularly need to load hundreds of various CSV (comma separated values) files with daily data from meteorological observations. For example, many of the measurements I use are made at the Reading University Atmospheric Observatory using the main datalogger, in addition to some of my own instruments. This data comes distributed across a number of different files for each day.

Most of my analysis is done in Python using the Pandas library for data processing. Pandas can easily read in CSV files with a built-in function, and it is well-suited for the two-dimensional data tables which I regularly use.

However, after a year or so of working directly with CSV files, I began to run up against some of the performance limitations of doing so.

Daily CSV files may be good for organizational purposes, but they are not the most efficient way to store and access large amounts of data. In particular, once I wanted to start studying many years’ worth of data at the same time, reading in each file every time I wanted to re-run my code began to slow the development process significantly. Also, the code that I had written to locate each file and read it in for a given range of dates was somewhat clunky and inflexible.

It was time to develop a new solution for accessing the met observations more quickly and easily.

Pandas has built-in functions for reading a variety of different formats, not just plain-text CSV files. I decided to constrain my choices for data formats to those that Pandas could read natively.

In addition, I wanted to build a system that would satisfy three primary goals:

  • Compatibility for long-term data storage
  • High speed
  • Simple programming interface

Compatibility is important, since I wanted to ensure that my data would continue to be readable to others (and myself in the future) without any specialized software that I had written. CSV is excellent for that purpose.

However, CSV was not a fast way to access the data. Ideally, the system I chose could store both numerical data and timestamps as floating point values rather than encoded text, for better performance.

Finally, I wanted to create a system that would be flexible and easy to use–ideally, something that would only require one or two lines of code to load in the data from a given instrument and date range, rather than the many complicated steps that had been required to search for and load the many files I had been using.

System Design

In the end, I settled on a rather complicated system that resulted in a very simple, reliable, and fast data stack that could be used to access my data.

At the base layer, all the data would be stored in the original CSV files. This is the format that most of the data comes in, and the few instruments that do not can easily be converted. CSV is a very common file format, which can be read easily by many software packages and will likely be useful far into the future, even when current software is too outdated to be run.

However, rather than directly accessing the CSV files, I import them occasionally into a SQLite database file. SQLite is a widely-used, open source software library which enables users to run a database from a single file, rather than a server (as opposed to many other popular database programs). The advantage over CSV files is that it is relatively fast. Data from an individual table can be accessed by a query specifying the start and end dates. This means that it is very easy to load in arbitrary timeseries of data.

However, for loading many years’ worth of high-resolution data, even SQLITE was not as fast as I wanted. Pandas is also capable of using a format called pickle. “Pickling” a dataframe outputs the dataframe from program memory to the disk as a file. This can be then be read back very quickly into a program at a later time, even for large files.

In my data access library, once a request is made for a given timeseries of data, that dataframe is cached to a pickle file. If the same request is made again shortly afterwards, rather than going back to the SQLite database, the data is loaded from the pickle file. For large datasets, this can reduce the loading time from nearly a minute to just a few seconds, which is very helpful when repeatedly debugging a program using the same data! The cache files can be relatively large, however, so they are automatically cleared out when the code runs if they have not been used for several days.

Finally, all of this functionality is available behind a simple library, which allows for accessing a dataset from any other Python code on my machine with just two lines, as shown below.

import foglib.db as fdb
fdb.load_db("dataset_name","start_datetime","end_datetime")

Conclusions

I have found this system to work very well for my purposes. It required a fair amount of development work, but the returns have been very beneficial.

By allowing me to access almost any of my data with just a few lines of code, I can now start new analyses with less time and code overhead. This means that I have more time and energy to spend answering science questions. And because it allows reading in large datasets so quickly, this means that I can rapidly debug my code without requiring me to wait as long while my code runs.

My particular solution may not be the ideal data-loading system for everyone’s needs. However, based on my experiences working on this program, I believe that time invested in enabling access to your data at the beginning of a PhD is time very well spent.

Investigating the use of early satellite data to test historical reconstructions of sea surface temperature

Thomas Hall1 Comment

Email: t.hall@pgr.reading.ac.uk

Observations of sea surface temperature (SST) form one of the key components of the climate record. There are a number of different in-situ based reconstructions of SST extending back over 150 years, but they are not truly independent of each other because the observations they are based on are largely the same (Berry et al., 2018). Datasets of SST retrieved from satellite radiometers exist for the 1980s onwards, providing an independent record of SST. Before this, SST reconstructions are based on sparse, ship-based measurements.

There were meteorological measurements being made from satellites in the 1960s and 70s, however, some of which can potentially be used to retrieve SST. My PhD focuses on investigating if we can retrieve SST from one of these early satellite instruments, to test the reliability of the SST reconstructions before the 1980s. This instrument is the Infrared Interferometer Spectrometer (IRIS), which made measurements of atmospheric emission spectra on-board the Nimbus 4 satellite from April 1970 to January 1971. IRIS had over 800 thermal infrared (IR) channels, covering the 400-1600 cm-1 spectral region. Figure 1 shows an example of two typical IRIS radiance spectra, with the main spectral absorption features labelled as well as the atmospheric window regions, which are the main spectral regions used for SST retrieval.

blog_fig1
Figure 1: Example of two typical IRIS radiance spectra; the main spectral absorption features are labelled as well as the atmospheric window regions.

Before using the IRIS data to retrieve SST, it was necessary to apply a series of quality assurance tests to filter out bad data. A few months into my PhD, work by Bantges et al. (2016) revealed evidence for a wavelength dependent cold bias of up to 2K in the data. A large part of my PhD was spent trying to quantify this bias. This was done by comparing clear-sky IRIS spectra with spectra simulated with a radiative transfer model. Unfortunately, this meant that the SSTs eventually retrieved from IRIS are not totally independent of the SST reconstructions as the simulations are based on reanalysis data forced by the HadISST2 reconstruction. Figure 2 compares our estimate of the IRIS spectral bias with globally averaged spectral differences between IRIS, the Interferometric Monitor for Greenhouse Gases (IMG) and the Infrared Atmospheric Sounding Instrument (IASI) from Bantges et al. (2016). This shows close agreement between our bias estimate and the IRIS-IMG and IRIS-IASI differences outside of the ozone spectral region, which is not relevant for SST retrieval.

It cannot just be assumed that the bias is the same for each IRIS measurement. Comparison of IRIS (bias-corrected using our initial bias estimate) with window channel data from the Temperature-Humidity Infrared Radiometer (THIR), also on-board Nimbus 4, reveals that the relative IRIS-THIR bias varies with window brightness temperature and orbit angle. The THIR, however, may have biases of its own, so these biases cannot be attributed to IRIS.

blog_fig2
Figure 2: Area-weighted global mean brightness temperature difference averaged over AMJ between IRIS (1970), IMG (1997) and IASI (2012) (black and blue lines) from Bantges et al. (2016), compared with our IRIS bias estimate, also area-weighted and averaged over AMJ (red line). The ozone absorption band is not used for SST retrieval, so is shaded grey.

The technique of optimal estimation was applied to retrieve SST from IRIS. This uses the observation-simulation differences together with information about the sensitivity of the simulations to the state of the atmosphere and ocean to estimate the SST. IR satellite retrievals of SST are usually performed in clear-sky conditions only. However, the low spatial resolution of IRIS means that very few cases are fully clear-sky. For this reason, we had to adapt the retrieval method to be tolerant of some cloud. This involves retrieving SST simultaneously with cloud fraction (CF). The retrieval method was then tested on partly cloudy (≤0.2 CF) IASI spectra made ‘IRIS-like’ by spatial averaging, spectral smoothing and simulating IRIS-like errors. The retrieved IRIS-like SSTs were validated against quality-controlled drifting buoy SSTs. This revealed latitudinal biases in the retrieved SSTs for the partly cloudy cases, not present in the SSTs for clear-sky cases.

SSTs were then retrieved for all IRIS cases with an expected CF ≤ 0.2. Figure 3 shows the difference between the gridded, monthly average IRIS SSTs and two of the SST reconstructions (HadSST3 and HadISST2) for July 1970. There are large, spatially correlated differences between the IRIS SSTs and reconstructions. We expect a latitudinal bias in the IRIS SSTs and some level of remaining bias in the IRIS spectra is likely, contributing to further SST bias. It is therefore likely that the differences in Figure 3 are mainly due to bias in the IRIS SSTs rather than the reconstructions.

iris_recon_07_70
Figure 3: Gridded IRIS-HadSST3 (left) and IRIS-HadISST2 (right) SST for July 1970. HadISST2 is a globally complete, interpolated dataset whereas HadSST3 is not globally complete.

Despite being unable to retrieve bias-free SST estimates from IRIS, my work has contributed towards better understanding the characteristics of IRIS. This ties in with a current project aiming to recover and assess the quality of data from a number of different historic satellite sensors, including IRIS, for assimilation in the next generation of climate reanalyses.

References

Bantges, R., H. Brindley, X. H. Chen, X. L. Huang, J. Harries, J. Murray (2016), On the detection of robust multi-decadal changes in the Earth’s Outgoing Longwave Radiation spectrum. J. Climate, 29, 4939-4947. https://doi.org/10.1175/JCLI-D-15-0713.1

Berry, D. I., G. K. Corlett, O. Embury, C. J. Merchant (2018), Stability assessment of the (A)ATSR Sea Surface Temperature climate dataset from the European Space Agency Climate Change Initiative. Remote Sens., 10, 126. https://doi.org/10.3390/rs10010126

Evaluating aerosol forecasts in London

Elliott Warren1 Comment

Email: e.l.warren@pgr.reading.ac.uk

Aerosols in urban areas can greatly impact visibility, radiation budgets and our health (Chen et al., 2015). Aerosols make up the liquid and solid particles in the air that, alongside noxious gases like nitrogen dioxide, are the pollution in cities that we often hear about on the news – breaking safety limits in cities across the globe from London to Beijing. Air quality researchers try to monitor and predict aerosols, to inform local councils so they can plan and reduce local emissions.

Figure 1: Smog over London (Evening Standard, 2016).

Recently, large numbers of LiDARs (Light Detection and Ranging) have been deployed across Europe, and elsewhere – in part to observe aerosols. They effectively shoot beams of light into the atmosphere, which reflect off atmospheric constituents like aerosols. From each beam, many measurements of reflectance are taken very quickly over time – and as light travels further with more time, an entire profile of reflectance can be constructed. As the penetration of light into the atmosphere decreases with distance, the reflected light is usually commonly called attenuated backscatter (β). In urban areas, measurements away from the surface like these are sorely needed (Barlow, 2014), so these instruments could be extremely useful. When it comes to predicting aerosols, numerical weather prediction (NWP) models are increasingly being considered as an option. However, the models themselves are very computationally expensive to run so they tend to only have a simple representation of aerosol. For example, for explicitly resolved aerosol, the Met Office UKV model (1.5 km) just has a dry mass of aerosol [kg kg-1] (Clark et al., 2008). That’s all. It gets transported around by the model dynamics, but any other aerosol characteristics, from size to number, need to be parameterised from the mass, to limit computation costs. However, how do we know if the estimates of aerosol from the model are actually correct? A direct comparison between NWP aerosol and β is not possible because fundamentally, they are different variables – so to bridge the gap, a forward operator is needed.

In my PhD I helped develop such a forward operator (aerFO, Warren et al., 2018). It’s a model that takes aerosol mass (and relative humidity) from NWP model output, and estimates what the attenuated backscatter would be as a result (βm). From this, βm could be directly compared to βo and the NWP aerosol output evaluated (e.g. see if the aerosol is too high or low). The aerFO was also made to be computationally cheap and flexible, so if you had more information than just the mass, the aerFO would be able to use it!

Among the aerFO’s several uses (Warren et al., 2018, n.d.), was the evaluation of NWP model output. Figure 2 shows the aerFO in action with a comparison between βm and observed attenuated backscatter (βo) measured at 905 nm from a ceilometer (a type of LiDAR) on 14th April 2015 at Marylebone Road in London. βm was far too high in the morning on this day. We found that the original scheme the UKV used to parameterise the urban surface effects in London was leading to a persistent cold bias in the morning. The cold bias would lead to a high relative humidity, so consequently the aerFO condensed more water than necessary, onto the aerosol particles as a result, causing them to swell up too much. As a result, bigger particles mean bigger βm and an overestimation. Not only was the relative humidity too high, the boundary layer in the NWP model was developing too late in the day as well. Normally, when the surface warms up enough, convection starts, which acts to mix aerosol up in the boundary layer and dilute it near the surface. However, the cold bias delayed this boundary layer development, so the aerosol concentration near the surface remained high for too long. More mass led to the aerFO parameterising larger sizes and total numbers of particles, so overestimated βm. This cold bias effect was reflected across several cases using the old scheme but was notably smaller for cases using a newer urban surface scheme called MORUSES (Met Office – Reading Urban Surface Exchange Scheme). One of the main aims for MORUSES was to improve the representation of energy transfer in urban areas, and at least to us it seemed like it was doing a better job!

Figure 2: Vertical profiles of attenuated backscatter [m−1 sr−1] (log scale) that are (a, g) observed (βo) with estimated mixing layer height (red crosses, Kotthaus and Grimmond,2018) and (b, h) forward modelled (βm) using the aerFO (section 2).(c, i) Attenuated backscatter difference (βm – βo) calculated using the hourly βm vertical profile and the vertical profile of βo nearest in time; (d, j) aerosol mass mixing ratio (m) [μg kg−1]; (e, k) relative humidity (RH) [%] and (f, l) air temperature (T) [°C] at MR on 14th April 2015.

References

Barlow, J.F., 2014. Progress in observing and modelling the urban boundary layer. Urban Clim. 10, 216–240. https://doi.org/10.1016/j.uclim.2014.03.011

Chen, C.H., Chan, C.C., Chen, B.Y., Cheng, T.J., Leon Guo, Y., 2015. Effects of particulate air pollution and ozone on lung function in non-asthmatic children. Environ. Res. 137, 40–48. https://doi.org/10.1016/j.envres.2014.11.021

Clark, P.A., Harcourt, S.A., Macpherson, B., Mathison, C.T., Cusack, S., Naylor, M., 2008. Prediction of visibility and aerosol within the operational Met Office Unified Model. I: Model formulation and variational assimilation. Q. J. R. Meteorol. Soc. 134, 1801–1816. https://doi.org/10.1002/qj.318

Warren, E., Charlton-Perez, C., Kotthaus, S., Lean, H., Ballard, S., Hopkin, E., Grimmond, S., 2018. Evaluation of forward-modelled attenuated backscatter using an urban ceilometer network in London under clear-sky conditions. Atmos. Environ. 191, 532–547. https://doi.org/10.1016/j.atmosenv.2018.04.045

Warren, E., Charlton-Perez, C., Kotthaus, S., Marenco, F., Ryder, C., Johnson, B., Lean, H., Ballard, S., Grimmond, S., n.d. Observed aerosol characteristics to improve forward-modelled attenuated backscatter. Atmos. Environ. Submitted


Understanding the Processes Involved in Electrifying Convective Clouds

James GilmoreLeave a comment

All clouds within our atmosphere are charged to some extent (Nicoll & Harrison, 2016), caused by the build-up of charge at the cloud edges caused by the charge travelling from the top of the atmosphere towards the surface. On the other hand, nearly all convective clouds are actively charged, caused by charge separation mechanisms that exchange charge between different sized hydrometeors within the cloud (Saunders, 1992). If these actively charged clouds can separate enough charge, then electrical breakdown can occur in the atmosphere and lightning can be initiated.

As lightning is a substantial hazard to both life and infrastructure, understanding the processes that cause a cloud to become charged are crucial for being able to forecast it. Even though most convective clouds within the UK are charged, most will never produce lightning. This puts us in a unique position of observing clouds which can and cannot produce lightning, providing a contrast of the cloud processes involved.

Cumliform_Electrification_Hypothesis
Figure 1: A conceptual image of a cloud showing the main processes that are thought to be involved in the electrification of a convective cloud. Based on current literature. See MacGorman & Rust (1998) for an overview.

Currently, lightning must be observed before the identification of a thunderstorm can be made, leaving zero lead time. As the first lightning strike can often be the most powerful, (especially in UK winter thunderstorms), improvements to understand the processes that charge a cloud are important to provide a thunderstorm lead time.

There are many processes involved in charging a cloud, a summary of the processes that are thought to have the greatest contribution can be found in Figure 1. The separation of charge through collisions is thought to be the dominant electrification mechanism which occurs in the ice phase of the cloud. The most successful charge separation occurs between growing ice hydrometeors of different sizes (Emersic & Saunders, 2010). The growth of the ice forms an outer shell of supercooled liquid water which contains negative charge. Collisions between different sized hydrometeors cause a net exchange of mass and charge, creating positively and negatively charged ice.

The liquid phase is crucial to maintaining a high moisture content in the ice phase of the cloud brought up by the updraught of the cloud. Once the electric field within the cloud is strong enough, the liquid drops can become polarised which can also help separate charge from collisions of different sized liquid hydrometeors in the later stages of convective cloud development.

Once the charge has been separated, the different polarities must be moved to different regions of the cloud to enhance the electric field. The most well-established method involves gravitational separation (Mason & Dash, 2000). Under the assumption that smaller hydrometeors will often contain negative charge and larger hydrometeors will contain positive charge, there is a distinct separation of hydrometeor sizes. After the formation of an updraught, the smaller hydrometeors can be lifted higher into the cloud overcoming the gravitational forces. Under the right conditions, the larger hydrometeors will be too heavy, and gravity will force the hydrometeors to remain lower in the cloud.

Field_Campaign_Instruments_Small
Figure 2: The Electrostatic Biral Thunderstorm Detector (BTD-300) (a), Electrostatic JCI 131 Field Mill (b), and 35 GHz “Copernicus” Radar installed at Chilbolton Observatory, UK. The Field Mill and BTD-300 were installed on 06/10/16 and 16/11/16 respectively.

To understand if any of the processes discussed in Figure 1 are evident within real-world convective clouds, a field campaign was set-up at Chilbolton Observatory (CO) to measure the properties of the cloud. Two electrical instruments were installed at CO, the Biral Thunderstorm Detector (BTD-300, Figure 2a) and the electrostatic Field Mill (FM, Figure 2b) which can measure the displacement current (jD) and the potential gradient (PG) respectively. The 35GHz “Copernicus” dopplerised radar was used to measure the properties of the cloud. Through a two-year field campaign, 653 convective clouds were identified over CO, with 524 clouds (80.2%) found to be charged and 129 clouds (19.8%) found to be uncharged.

To understand the importance of hydrometeor size for charging a cloud, the 95th percentile of the radar reflectivity (Z) was used. The Z is strongly related to the diameter (sixth power) and number concentration of hydrometeors (first power). Figure 3 shows a boxplot of all 653 clouds, classified by the cloud charge as measured at the surface. For each cloud, the liquid and ice regions were separated (purple and blue boxplots respectively) to highlight the dominant region for charge separation.

Z_Grouped_BoxPlot_LinearScale_v22
Figure 3: A boxplot of the 95th percentile reflectivity for 635 identified clouds (black box) classified into no charge, small charge and large charge groups. The cloud was also decoupled into the liquid (purple box) and ice (blue box) phases. The box-plot shows the mean (purple bar), median (red bar) and upper and lower quartiles (upper and lower limits of black box).

For all phases of the cloud, there was a substantial increase in the reflectivity for all regions of the cloud, especially the liquid phase. This suggests that the size of the hydrometeors in both the ice and liquid phase are indeed important for charging a cloud.

In the remainder of my PhD, the relative importance of each process discussed in Figure 1 will be addressed to try and decouple each process. Further to these observations made at the surface, ten radiosonde flights will be made (from the Reading University Atmospheric Observatory) inside convectively charged clouds. Measurements of the charge, optical thickness, amount of supercooled liquid water and turbulence will be used to increase the robustness of the results presented in the previous works.

Email: james.gilmore@pgr.reading.ac.uk

References

Bouniol, D., Illingworth, A. J. & Hogan, R. J., 2003. Deriving turbulent kinetic energy dissipation rate within clouds using ground-based 94 GHz radar. Seattle, 31st International Conference on Radar Meteorology.

Emersic, C. & Saunders, C. P., 2010. Further laboratory investigations into the Relative Diffusional Growth Rate theory of thunderstorm electrification. Atmos. Res., Volume 98, pp. 327-340.

MacGorman, D. R. & Rust, D. W., 1998. The Electrical Nature of Storms. 1st ed. New York: Oxford University Press.

Mason, B. L. & Dash, J. G., 2000. Charge and mass transfer in ice–ice collisions: Experimental observations of a mechanism in thunderstorm electrification. J. Geophys. Res., Volume 105, pp. 10185-10192.

Nicoll, K. A. & Harrison, R. G., 2016. Stratiform cloud electrification: comparison of theory with multiple in-cloud measurements. Q.J.R. Meteorol. Soc., Volume 142, pp. 2679-2691.

Renzo, M. D. & Urzay, J., 2018. Aerodynamic generation of electric fields in turbulence laden with charged inertial particles. Nat. Comms., 9(1676).

Saunders, C. P. R., 1992. A Review of Thunderstorm Electrification Processes. J. App. Meteo., Volume 32, pp. 642-655.

Baroclinic and Barotropic Annular Modes of Variability

Lina Boljka1 Comment

Email: l.boljka@pgr.reading.ac.uk

Modes of variability are climatological features that have global effects on regional climate and weather. They are identified through spatial structures and the timeseries associated with them (so-called EOF/PC analysis, which finds the largest variability of a given atmospheric field). Examples of modes of variability include El Niño Southern Oscillation, Madden-Julian Oscillation, North Atlantic Oscillation, Annular modes, etc. The latter are named after the “annulus” (a region bounded by two concentric circles) as they occur in the Earth’s midlatitudes (a band of atmosphere bounded by the polar and tropical regions, Fig. 1), and are the most important modes of midlatitude variability, generally representing 20-30% of the variability in a field.

Southern_Hemi_Antarctica
Figure 1: Southern Hemisphere midlatitudes (red concentric circles) as annulus, region where annular modes have the largest impacts. Source.

We know two types of annular modes: baroclinic (based on eddy kinetic energy, a proxy for eddy activity and an indicator of storm-track intensity) and barotropic (based on zonal mean zonal wind, representing the north-south shifts of the jet stream) (Fig. 2). The latter are usually referred to as Southern (SAM or Antarctic Oscillation) or Northern (NAM or Arctic Oscillation) Annular Mode (depending on the hemisphere), have generally quasi-barotropic (uniform) vertical structure, and impact the temperature variations, sea-ice distribution, and storm paths in both hemispheres with timescales of about 10 days. The former are referred to as BAM (baroclinic annular mode) and exhibit strong vertical structure associated with strong vertical wind shear (baroclinicity), and their impacts are yet to be determined (e.g. Thompson and Barnes 2014, Marshall et al. 2017). These two modes of variability are linked to the key processes of the midlatitude tropospheric dynamics that are involved in the growth (baroclinic processes) and decay (barotropic processes) of midlatitude storms. The growth stage of the midlatitude storms is conventionally associated with increase in eddy kinetic energy (EKE) and the decay stage with decrease in EKE.

ThompsonWoodworth_Fig2a_SAM_2f_BAM(1)
Figure 2: Barotropic annular mode (right), based on zonal wind (contours), associated with eddy momentum flux (shading); Baroclinic annular mode (left), based on eddy kinetic energy (contours), associated with eddy heat flux (shading). Source: Thompson and Woodworth (2014).

However, recent observational studies (e.g. Thompson and Woodworth 2014) have suggested decoupling of baroclinic and barotropic components of atmospheric variability in the Southern Hemisphere (i.e. no correlation between the BAM and SAM) and a simpler formulation of the EKE budget that only depends on eddy heat fluxes and BAM (Thompson et al. 2017). Using cross-spectrum analysis, we empirically test the validity of the suggested relationship between EKE and heat flux at different timescales (Boljka et al. 2018). Two different relationships are identified in Fig. 3: 1) a regime where EKE and eddy heat flux relationship holds well (periods longer than 10 days; intermediate timescale); and 2) a regime where this relationship breaks down (periods shorter than 10 days; synoptic timescale). For the relationship to hold (by construction), the imaginary part of the cross-spectrum must follow the angular frequency line and the real part must be constant. This is only true at the intermediate timescales. Hence, the suggested decoupling of baroclinic and barotropic components found in Thompson and Woodworth (2014) only works at intermediate timescales. This is consistent with our theoretical model (Boljka and Shepherd 2018), which predicts decoupling under synoptic temporal and spatial averaging. At synoptic timescales, processes such as barotropic momentum fluxes (closely related to the latitudinal shifts in the jet stream) contribute to the variability in EKE. This is consistent with the dynamics of storms that occur on timescales shorter than 10 days (e.g. Simmons and Hoskins 1978). This is further discussed in Boljka et al. (2018).

EKE_hflux_cross_spectrum_blog
Figure 3: Imaginary (black solid line) and Real (grey solid line) parts of cross-spectrum between EKE and eddy heat flux. Black dashed line shows the angular frequency (if the tested relationship holds, the imaginary part of cross-spectrum follows this line), the red line distinguishes between the two frequency regimes discussed in text. Source: Boljka et al. (2018).

References

Boljka, L., and T. G. Shepherd, 2018: A multiscale asymptotic theory of extratropical wave, mean-flow interaction. J. Atmos. Sci., in press.

Boljka, L., T. G. Shepherd, and M. Blackburn, 2018: On the coupling between barotropic and baroclinic modes of extratropical atmospheric variability. J. Atmos. Sci., in review.

Marshall, G. J., D. W. J. Thompson, and M. R. van den Broeke, 2017: The signature of Southern Hemisphere atmospheric circulation patterns in Antarctic precipitation. Geophys. Res. Lett., 44, 11,580–11,589.

Simmons, A. J., and B. J. Hoskins, 1978: The life cycles of some nonlinear baroclinic waves. J. Atmos. Sci., 35, 414–432.

Thompson, D. W. J., and E. A. Barnes, 2014: Periodic variability in the large-scale Southern Hemisphere atmospheric circulation. Science, 343, 641–645.

Thompson, D. W. J., B. R. Crow, and E. A. Barnes, 2017: Intraseasonal periodicity in the Southern Hemisphere circulation on regional spatial scales. J. Atmos. Sci., 74, 865–877.

Thompson, D. W. J., and J. D. Woodworth, 2014: Barotropic and baroclinic annular variability in the Southern Hemisphere. J. Atmos. Sci., 71, 1480–1493.

Trouble in paradise: Climate change, extreme weather and wildlife conservation on a tropical island.

Social Metwork GuestLeave a comment

Joseph Taylor, NERC SCEARNIO DTP student. Zoological Society of London.

Email: J.Taylor5@pgr.reading.ac.uk

Projecting the impacts of climate change on biodiversity is important for informing

Mauritius Kestrel by Joe Taylor
Male Mauritius kestrel (Falco punctatus) in the Bambous Mountains, eastern Mauritius. Photo by Joe Taylor.

mitigation and adaptation strategies. There are many studies that project climate change impacts on biodiversity; however, changes in the occurrence of extreme weather events are often omitted, usually because of insufficient understanding of their ecological impacts. Yet, changes in the frequency and intensity of extreme weather events may pose a greater threat to ecosystems than changes in average weather regimes (Jentsch and Beierkuhnlein 2008). Island species are expected to be particularly vulnerable to climate change pressures, owing to their inherently limited distribution, population size and genetic diversity, and because of existing impacts from human activities, including habitat destruction and the introduction of non-native species (e.g. Fordham and Brook 2010).

Mauritius is an icon both of species extinction and the successful recovery of threatened species. However, the achievements made through dedicated conservation work and the investment of substantial resources may be jeopardised by future climate change. Conservation programmes in Mauritius have involved the collection of extensive data on individual animals, creating detailed longitudinal datasets. These provide the opportunity to conduct in-depth analyses into the factors that drive population trends.

My study focuses on the demographic impacts of weather conditions, including extreme events, on three globally threatened bird species that are endemic to Mauritius. I extended previous research into weather impacts on the Mauritius kestrel (Falco punctatus), and applied similar methods to the echo parakeet (Psittacula eques) and Mauritius fody (Foudia rubra). The kestrel and parakeet were both nearly lost entirely in the 1970s and 1980s respectively, having suffered severe population bottlenecks, but all three species have benefitted from successful recovery programmes. I analysed breeding success using generalised linear mixed models and analysed survival probability using capture-mark-recapture models. Established weather indices were adapted for use in this study, including indices to quantify extreme rainfall, droughts and tropical cyclone activity. Trends in weather indices at key conservation sites were also analysed.

The results for the Mauritius kestrel add to a body of evidence showing that precipitation is an important limiting factor in its demography and population dynamics. The focal population in the Bambous Mountains of eastern Mauritius occupies an area in which rainfall is increasing. This trend could have implications for the population, as my analyses provide evidence that heavy rainfall during the brood phase of nests reduces breeding success, and that prolonged spells of rain in the cyclone season negatively impact the survival of juveniles. This probably occurs through reductions in hunting efficiency, time available for hunting and prey availability, so that kestrels are unable to capture enough prey to sustain themselves and feed their young (Nicoll et al. 2003, Senapathi et al. 2011). Exposure to heavy and prolonged rainfall could also be a direct cause of mortality through hypothermia, especially for chicks if nests are flooded (Senapathi et al. 2011). Future management of this species may need to incorporate strategies to mitigate the impacts of increasing rainfall.

References:

Fordham, D. A. and Brook, B. W. (2010) Why tropical island endemics are acutely susceptible to global change. Biodiversity and Conservation 19(2): 329‒342.

Jentsch, A. and Beierkuhnlein, C. (2008) Research frontiers in climate change: Effects of extreme meteorological events on ecosystems. Comptes Rendus Geoscience 340: 621‒628.

Nicoll, M. A. C., Jones, C. G. and Norris, K. (2003) Declining survival rates in a reintroduced population of the Mauritius kestrel: evidence for non-linear density dependence and environmental stochasticity. Journal of Animal Ecology 72: 917‒926.

Senapathi, D., Nicoll, M. A. C., Teplitsky, C., Jones, C. G. and Norris, K. (2011) Climate change and the risks associated with delayed breeding in a tropical wild bird population. Proceedings of the Royal Society B 278: 3184‒3190.

Climate model systematic biases in the Maritime Continent

Ying Ying TohLeave a comment

Email: y.y.toh@pgr.reading.ac.uk

The Maritime Continent commonly refers to the groups of islands of Indonesia, Borneo, New Guinea and the surrounding seas in the literature. My study area covers the Maritime Continent domain from 20°S to 20°N and 80°E to 160°E as shown in Figure 1. This includes Indonesia, Malaysia, Brunei, Singapore, Philippines, Papua New Guinea, Solomon islands, northern Australia and parts of mainland Southeast Asia including Thailand, Laos, Cambodia, Vietnam and Myanmar.

subsetF1
Figure 1: JJA precipitation (mm/day) and 850 hPa wind (m s−1) for (a) GPCP and ERA-interim, (b) MMM biases and (c)–(j) AMIP biases for 1979–2008 over the Maritime Continent region (20°S–20ºN, 80°E–160ºE). Third panel shows the Maritime Continent domain and land-sea mask

The ability of climate model to simulate the mean climate and climate variability over the Maritime Continent remains a modelling challenge (Jourdain et al. 2013). Our study examines the fidelity of Coupled Model Intercomparison Project phase 5 (CMIP5) models at simulating mean climate over the Maritime Continent. We find that there is a considerable spread in the performance of the Atmospheric Model Intercomparison Project (AMIP) models in reproducing the seasonal mean climate and annual cycle over the Maritime Continent region. The multi-model mean (MMM) (Figure 1b) JJA precipitation and 850hPa wind biases with respect to observations (Figure 1a) are small compared to individual model biases (Figure 1c-j) over the Maritime Continent. Figure 1 shows only a subset of Fig. 2 from Toh et al. (2017), for the full figure and paper please click here.

We also investigate the model characteristics that may be potential sources of bias. We find that AMIP model performance is largely unrelated to model horizontal resolution. Instead, a model’s local Maritime Continent biases are somewhat related to its biases in the local Hadley circulation and global monsoon.

cluster2
Figure 2: Latitude-time plot of precipitation zonally averaged between 80°E and 160°E for (a) GPCP, (b) Cluster I and (c) Cluster II. White dashed line shows the position of the maximum precipitation each month. Precipitation biases with respect to GPCP for (d) Cluster I and (e) Cluster II.

To characterize model systematic biases in the AMIP runs and determine if these biases are related to common factors elsewhere in the tropics, we performed cluster analysis on Maritime Continent annual cycle precipitation. Our analysis resulted in two distinct clusters. Cluster I (Figure 2b,d) is able to reproduce the observed seasonal migration of Maritime Continent precipitation, but it overestimates the precipitation, especially during the JJA and SON seasons. Cluster II (Figure 2c,e) simulate weaker seasonal migration of Intertropical Convergence Zone (ITCZ) than observed, and the maximum rainfall position stays closer to the equator throughout the year. Tropics-wide properties of clusters also demonstrate a connection between errors at regional scale of the Maritime Continent and errors at large scale circulation and global monsoon.

On the other hand, comparison with coupled models showed that air-sea coupling yielded complex impacts on Maritime Continent precipitation biases. One of the outstanding problems in the coupled CMIP5 models is the sea surface temperature (SST) biases in tropical ocean basins. Our study highlighted central Pacific and western Indian Oceans as the key regions which exhibit the most surface temperature correlation with Maritime Continent mean state precipitation in the coupled CMIP5 models. Future work will investigate the impact of SST perturbations in these two regions on Maritime Continent precipitation using Atmospheric General Circulation Model (AGCM) sensitivity experiments.

 

 

References:

Jourdain N.C., Gupta A.S., Taschetto A.S., Ummenhofer C.C., Moise A.F., Ashok K. (2013) The Indo-Australian monsoon and its relationship to ENSO and IOD in reanalysis data and the CMIP3/CMIP5 simulations. Climate Dynamics. 41(11–12):3073–3102

Toh, Y.Y., Turner, A.G., Johnson, S.J., & Holloway, C.E. (2017). Maritime Continent seasonal climate biases in AMIP experiments of the CMIP5 multimodel ensemble. Climate Dynamics. doi: 10.1007/s00382-017-3641-x