School of Agriculture, Food and Ecosystem Sciences - Theses
Permanent URI for this collection
Search Results
1 - 1 of 1
-
ItemEffect of green roof design configuration and climate on rainfall retention, evapotranspiration, and plant drought stress( 2023-04)Urbanisation significantly alters the hydrological cycle through creation of impervious surfaces and removal of vegetation. Besides creating large volumes of stormwater runoff which degrades urban stream ecosystems and water sources, impervious surfaces reduce infiltration of rainfall and therefore water availability for remaining vegetation. Sustainable and resilient stormwater management techniques are required to mitigate the impacts of stormwater runoff and compensate for the loss of vegetation in cities. Green roofs are a promising green infrastructure technology with the potential to deliver these ecosystem services, however, gaps in our understanding of how they should be designed is preventing widespread uptake. Green roof substrates are typically shallow due to building weight-loading restrictions and therefore have limited water storage for reducing runoff and sustaining vegetation. Species typically used on green roofs are limited to those with the ability to survive harsh environmental conditions, such as succulent species. However, selecting plants with high water use, combined with drought resistance, such as non-succulent species, and planting at higher density in deeper substrates may improve green roof rainfall retention without substantially increasing plant drought stress. Where substrate depth cannot be increased, mimicking processes we observe in natural systems, such as redirecting rainfall towards vegetation in semi-arid banded systems, has significant potential to both increase rainfall retention and reduce plant drought stress. Climate is the primary driver of green roof performance, specifically the supply of and demand for water from the atmosphere, which in turn determines evapotranspiration, rainfall retention and plant drought stress. Ideally, by understanding the interaction of substrate depth, plant density (water use) and redistribution of water resources on green roofs, it should be possible to determine the most suitable green roof configuration for different climates, and therefore remove the key barrier to widespread green roof installation. In this thesis, I aimed to achieve this understanding through a combination of controlled experiments and water balance modelling. Firstly, a glasshouse experiment was used to understand how increasing substrate depth and plant density, as well as their interaction, impacted plant water use, drought stress and rainfall retention. Due to COVID-19 disruptions, only the first well-watered phase could be completed, after which the experiment had to be terminated without measuring water use and drought stress under water-deficit. Therefore, I used an established green roof water balance model to simulate performance under water deficit conditions. Pre-existing functions describing the plant species’ drought response were combined with well-watered plant crop factors (Kc) calculated from ET measured during the experiment, to estimate rainfall retention and the incidence of plant drought stress. Contrary to my initial hypotheses, increasing plant density did not result in a proportional increase in plant water use, even when substrate depth was doubled. Importantly, this indicated little gain in retention performance by increasing plant density. Using a water balance model to extend these findings to include performance under water deficit, I showed that rainfall retention was very high, regardless of substrate depth and plant density, as plants in the glasshouse had a very high crop factor and therefore rates of water use. With a high crop factor, all treatments from the glasshouse simulated in the model depleted the substrate water quickly, resulting in greater retention, but also significant plant drought stress. In this model, the indicator of drought stress in plants on each day of the rainfall simulation was when the depth of water (millimeter) in the substrate at the end of any given day reached zero. Importantly, increasing substrate depth showed no significant benefit to either rainfall retention or plant drought stress. Overall, planting in shallower substrates at a lower density optimised green roof performance when measured and simulated in a temperate climate. Secondly, a rainfall simulation experiment using green roof modules was conducted to understand the effect of plant density and redistribution of rainfall (runoff zones) on rainfall retention, plant water use and drought stress. In this experiment, drought stress in plants was indicated by midday leaf water potential (MegaPascal). Again, planting at a lower density (10 plants per module, approximately half the module area planted) achieved high rainfall retention and most importantly, plants experienced lower drought stress than fully-planted modules (18 plants per module). Furthermore, using runoff zones to direct rainfall towards plants also reduced plant drought stress. However, the runoff zones also created preferential flow pathways and shaded the substrate surface, both of which were the likely cause of lower rainfall retention, despite the observed reduction in plant drought stress. Although, reducing plant density showed the most effective way of achieving high rainfall retention and lower drought stress, there are other ways of increasing substrates' water retention and holding capacity, such as using water retention additives to increase water available for more densely planted green roofs which would improve the ecological, environmental, and social benefits of a green roofs. While redirecting rainfall showed promising approach, further work is required to improve their design to find the optimal method to redirect more water to plants and improve the coverage of plants on green roofs. Finally, using results from both experiments, I developed and validated a new green roof water balance model to simulate long-term green roof rainfall retention and plant performance in two contrasting climates (temperate vs semi-arid climates). Green roofs showed high rainfall retention in both temperate and semi-arid climates, regardless of substrate depth, plant density and presence/ absence of runoff zones. Even unplanted roofs showed high retention in both climates, showing that evaporation is the major component of evapotranspiration and therefore a primary driver of rainfall retention. In this experiment, midday leaf water potential was used to indicate the maximum water stress experienced by plants during the day and therefore plant drought stress. Therefore, I modified the water balance model used in the first experiment, by using the relationship between midday leaf water potential (MegaPascal) and substrate water content (S) to estimate plant drought stress. As expected, green roofs in semi-arid climates had significantly greater plant drought stress (more negative water potential) as compared with those in temperate climates, with no observed benefit in rainfall retention, despite increased substrate depth. Hence, a substrate depth of 150 millimeter could achieve optimal retention in both temperate and semi-arid climates. Increasing substrate depth, plant density and the use of runoff zones was less important for improving rainfall retention than climate. The modeled results also highlight that an unplanted roof is equally good for stormwater management alone, as it can achieve similar rates of rainfall retention as compared to a planted roof. However, keeping in mind the ecological, environmental, and social benefits of vegetated green roofs with good plant coverage, it is not recommended that practitioners install non-vegetated green roofs. Overall, the results showed that green roofs perform very well for rainfall retention, in both temperate climates with a large proportion of small rainfall events, and semi-arid climates with an annually low rainfall depth. However, only one plant species was evaluated in my thesis, which would have impacted on these results as the plants were high-water using species that could effectively dry out substrates after rainfall and also tolerate drought stress in dry substrates. The design of green roofs for good plant coverage and survival in semi-arid climates is likely to be more challenging than constructing in temperate climates in real conditions due to the risk of plant death where drought periods are more severe and prolonged. In this case, it is likely to be preferrable to plant low water using succulent species. In both temperate and semi-arid climates, green roof substrate depth did not necessarily need to be deeper than 150 mm, as the increase in rainfall retention was minimal and plant drought stress could not alleviate beyond this depth. Installing runoff zones is a promising approach to changing how water is distributed on green roofs and have the capacity to reduce plant drought stress. However, they can also promote preferential flow pathways minimising the water storage capacity of substrates and decreasing the evaporation from surface of substrates underneath their structure and therefore, reducing rainfall retention. Hence, reconsideration the design for such runoff zones could be a potential avenue of future research. In the end, the practical output of this research suggests that green roofs with lower plant density such as 1 plant per 0.1 square meter and substrate depth such as 150 millimeter can effectively retain rainfall in temperate and semi-arid climates. While in temperate climates, higher water using plant species can be used, it is recommended that in semi-arid climates, green roofs are planted with low water using species such as succulents to support vegetation cover on green roofs. This means that even when green roof plant cover reduces over time, green roofs will still have high performance for rainfall retention.