Fabio Veronesi

Landscape scale estimation of soil carbon stock using 3D modelling

Soil C is the largest pool of carbon in the terrestrial biosphere, and yet the processes of C accumulation, transformation and loss are poorly accounted for. This, in part, is due to the fact that soil C is not uniformly distributed through the soil depth profile and most current landscape level predictions of C do not adequately account the vertical distribution of soil C. In this study, we apply a method based on simple soil specific depth functions to map the soil C stock in three-dimensions at landscape scale. We used soil C and bulk density data from the Soil Survey for England and Wales to map an area in the West Midlands region of approximately 13,948 km(2). We applied a method which describes the variation through the soil profile and interpolates this across the landscape using well established soil drivers such as relief, land cover and geology. The results indicate that this mapping method can effectively reproduce the observed variation in the soil profiles samples. The mapping results were validated using cross validation and an independent validation. The cross-validation resulted in an R(2) of 36% for soil C and 44% for BULKD. These results are generally in line with previous validated studies. In addition, an independent validation was undertaken, comparing the predictions against the National Soil Inventory (NSI) dataset. The majority of the residuals of this validation are between ± 5% of soil C. This indicates high level of accuracy in replicating topsoil values. In addition, the results were compared to a previous study estimating the carbon stock of the UK. We discuss the implications of our results within the context of soil C loss factors such as erosion and the impact on regional C process models.

A GIS tool to increase the visual quality of relief shading by automatically changing the light direction

Shaded relief representations were traditionally produced manually by specifically trained cartographers. This was however a labour-intensive and time-consuming task, which gave rise to numerous attempts of automation.Nowadays, many GIS applications implement hillshading using an oblique light source. This has standardised the method, providing a simple way to obtain reliable and consistent results. Its visual quality is however well below the standards of manual shading, where multiple light sources are employed to achieve a superior visual quality.In this work we present a GIS tool to enhance the visual quality of hillshading. We developed a technique to illuminate the landscape from two different angles and correct the tone according to either elevation or slope. With this method the quality of the shaded relief is superior to the standard method, but its level of automation and standardisation guarantees consistent and reproducible results. This method has been embedded into an ArcGIS toolbox.

Assessing Accuracy and Geographical Transferability of Machine Learning Algorithms for Wind Speed Modelling

Machine learning is very popular in the environmental modelling community and has recently been demonstrated to be a useful tool for wind resource assessment as well. Despite the popularity of wind resource assessment, research in the field of machine learning for this purpose is in its infancy. Only few algorithms have been tested and only for specific areas, making it difficult to draw any conclusions in regards to the best wind estimation method at the global scale. In this study, we compared several machine learning algorithms with validation techniques specifically employed to not only assess their accuracy but also their transferability. In particular, we tested cross-validation techniques designed to test the accuracy of the estimation in the context of autocorrelation. This way we performed a benchmarking experiment that should provide end users with practical rules of application for each algorithm. We tested three families of popular algorithms, namely linear models, decision trees and support vector machines; each was tested using wind mean speed data and several environmental covariates as predictors. The results demonstrated that no single algorithm could consistently be used to estimate wind globally, even though decision-tree based methods seemed to be often the best estimators.

Statistical Learning Approach for Wind Speed Distribution Mapping: The UK as a Case Study

Wind resource assessment is fundamental when selecting a site for wind energy projects. Wind speed is influenced by a plethora of environmental factors and understanding its spatial variability is key for determining the economic viability of a site. Deterministic estimation methods, which are based on physics, represent the industry standard in wind-speed mapping. Over the years, these methods have proven capable of estimating wind speed with a relatively high accuracy. However measuring stations, which provide the starting data for all wind speed estimations, are often located at a distance from each other, in some cases, tens of kilometres or more. This adds an unavoidable level of uncertainty to the estimates, which deterministic methods fail to take into account. For this reason, even though there are ways of determining the overall uncertainty of the estimation, e.g. cross-validation, deterministic methods do not provide means of assessing the site-specific uncertainty. This paper introduces a statistical method for estimating wind speed, based on spatial statistics. In particular, we present a statistical learning approach, based on ensembles of regression trees, to estimate both the wind distribution in specific locations and to assess the site-specific uncertainty.

Changing the Light Azimuth in Shaded Relief Representation by Clustering Aspect

Abstract Manual shading, traditionally produced manually by specifically trained cartographers, is still considered superior to automatic methods, particularly for mountainous landscapes. However, manual shading is time-consuming and its results depend on the cartographer and as such difficult to replicate consistently. For this reason there is a need to create an automatic method to standardize its results. A crucial aspect of manual shading is the continuous change of light direction (azimuth) and angle (zenith) in order to better highlight discrete landforms. Automatic hillshading algorithms, widely available in many geographic information systems (GIS) applications, do not provide this feature. This may cause the resulting shaded relief to appear flat in some areas, particularly in areas where the light source is parallel to the mountain ridge. In this work we present a GIS tool to enhance the visual quality of hillshading. We developed a technique based on clustering aspect to provide a seamless change of lighting throughout the scene. We also provide tools to change the light zenith according to either elevation or slope. This way the cartographer has more room for customizing the shaded relief representation. Moreover, the method is completely automatic and this guarantees consistent and reproducible results. This method has been embedded into an ArcGIS toolbox.

Generation and Validation of Spatial Distribution of Hourly Wind Speed Time-Series using Machine Learning

Wind resource assessment is a key aspect of wind farm planning since it allows to estimate the long term electricity production. Moreover, wind speed time-series at high resolution are helpful to estimate the temporal changes of the electricity generation and indispensable to design stand-alone systems, which are affected by the mismatch of supply and demand. In this work, we present a new generalized statistical methodology to generate the spatial distribution of wind speed time-series, using Switzerland as a case study. This research is based upon a machine learning model and demonstrates that statistical wind resource assessment can successfully be used for estimating wind speed time-series. In fact, this method is able to obtain reliable wind speed estimates and propagate all the sources of uncertainty (from the measurements to the mapping process) in an efficient way, i.e. minimizing computational time and load. This allows not only an accurate estimation, but the creation of precise confidence intervals to map the stochasticity of the wind resource for a particular site. The validation shows that machine learning can minimize the bias of the wind speed hourly estimates. Moreover, for each mapped location this method delivers not only the mean wind speed, but also its confidence interval, which are crucial data for planners.

Comparison of hourly and daily wind speed observations for the computation of Weibull parameters and power output

In this paper we compared Weibull distributions fitted using the hourly wind speed observations measured by weather stations of the UK Meteorological Office and daily averages measured by the National Oceanographic and Atmospheric Administration in the same locations. The measurements collected over 5 years (2009-2013) have been analysed. The shape and scale parameters and the mean wind speed have been compared. Finally, we considered a 3 MW wind turbine with 100m hub height to estimate and compare the long term electricity generation using the two datasets. The results show that, using daily averages, the mean wind speed values do not significantly vary. However, its standard deviation varies substantially, which is due to the smoothing of extreme wind measurements in the daily average dataset. When performing the long term estimate of the full-load hours and the electricity generation, the results show that error margins are low when using daily averages.

Evaluation of River Water Quality: A Case Study of the Lea Navigation (NE London)

The Lea Navigation in the north-east of London, a canalized reach of the River Lea, is affected by episodes of very low levels of dissolved oxygen. The problem was detected by the Environment Agency from the confluence with Pymmes Brook (which receives the final effluent of Deephams sewage treatment works) to the Olympic site (Marshgate Lane, Stratford). In this study, possible causes and sources of the poor water quality in the Lea Navigation were investigated using algal bioassays and detailed spatial seasonal mapping of the physico-chemical parameters collected in situ. Results showed chronic pollution and identified polar compounds in the river water and high bacterial concentrations as possible causes of low dissolved oxygen levels. This study confirmed the negative impact of Deephams sewage treatment works (via Pymmes Brook) on the water quality of the Lea. However, whilst the Environment Agency had previously focused on the pollution created by the sewage treatment works, results showed evidence of other sources of pollution; in particular, Stonebridge Brook was identified as an uncontrolled source of pollution and untreated wastewater. This study demonstrates the value of conducting combined methodologies and detailed monitoring. Other possible sources include Old Moselle Brook, diffuse pollution from surface run-off, boat discharges and other undetected drainage misconnections.