May 30, 2021 Amsterdam
JULIETTE
ROOS
-
11860111
CARBON
STORAGE
IN
HEDGE
BIOMASS
A CASE STUDY OF HEDGEROWS IN THE AGRICULTURAL
LANDSCAPE
Bsc Future Planet Studies & major Future Earth
University of Amsterdam
dhr. dr. D. A.C. (Arie Christoffel)
Abstract
Agricultural lands are solely managed to maximize the yield of provisioning ecosystem services, like food (Power, 2010). However, this comes at the expense of regulating ecosystem services, including carbon storage. Carbon capture and storage (CCS) is critical to cut atmospheric concentrations of CO2 to combat climate change. One way to increase the carbon storage in the agricultural landscape is by planting more hedgerows. However, research on carbon storage in hedge biomass is lacking (Axe, Grange, & Conway, 2017). Thankfully, airborne Lidar is becoming more widely recognized as a remote sensing tool capable of adequately measuring carbon storage on a broad scale when used in conjunction with an empirical model (Popescu, 2007a). Nevertheless, scientific literature on an empirical model is lacking (Axe et al., 2017).
This research aims to quantify the carbon storage in the woody biomass of hedgerows and investigate the added value of Lidar data for estimating large scale carbon storage in hedgerows. As a case study, we look at a trimmed hawthorn hedgerow on farm Hofstede Rhijnauwen in the Netherlands. Carbon storage per square meter hedge was calculated by the use of a new derived method for quantification of hedgerow biomass in addition to 308 measurements of the sampled hedgerow. Also, ALS height metrics were checked on accuracy with the help of MLS 3d-plots in RStudio. Finally, using ALS height metrics and field data, the total carbon storage of the study area was determined.
The results show that a trimmed hawthorn hedgerow with a height of 1.38-meter stores 11.66 kg carbon m-2 hedge. Furthermore, without near vegetation, AHN height metrics are correct. At last, the hedgerows of Hofstede Rhijnauwen store 15080 kg of carbon. Based on the results of this research, it can be concluded that reintroducing hedgerows is a potential solution as a natural intervention for increasing carbon storage in the agricultural landscape. The result of this research could be used to encourage the reintroduction of hedgerows as it demonstrated a new method for quantifying hedgerow biomass in combination with Lidar data.
Content list
Introduction 3
Background & relevance 3
Research aim & questions 3
Methods and data 5
Data 5
Field- and lab work data 5
AHN & ZEB-REVO data 5
Methods 6
Field- and lab work data 6
AHN & ZEB-REVO data 6
Results 9
Carbon storage m-2 hedge 9
Accuracy of LiDAR data 9
AHN metrics 9
ZEB-REVO and AHN; Canopy Height Model and 3D-plot 10
Total carbon Storage in the hedgerows of Hofsted Rhijnauwen 10
Discussion 12
Interpretation of the results 12
Methodological discussion 12 Outlook 13 Conclusion 14 References 15 Acknowledgements 18 Appendices 19
Appendix A: RStudio code 19
Appendix B: Field data on hedgerow height 20
Introduction
Background & relevance
Agricultural lands are primarily managed to maximize the yield of provisioning ecosystem services, like food and fibre (Power, 2010). Nevertheless, this comes at the expense of the provisioning of regulating ecosystem services. Maintaining the quality of air, disease control, and carbon storage are some of the regulating services provided by ecosystems. These are often invisible and therefore taken for granted. Yet, when damaged, the resulting losses are substantial (FAO, 2021). Zooming in on the regulating service of carbon sequestration, carbon dioxide (CO2) is the most important anthropogenic greenhouse gas (GHG) (Joos et al., 2001). With rising concern over climate change, global efforts are made to reduce emissions and increase carbon storage. Hereby, carbon capture and storage (CCS) is critical to cut atmospheric concentrations of CO2.
One way to increase carbon storage in agricultural lands is by planting hedgerows. Over the past decades, the implementation of hedgerows in the agricultural landscape gained attention (Holden et al., 2019) while the demand for most ecosystem services is increasing (Assessment, 2005). However, this is mainly focused on other ecosystem services, such as improved crop productivity, reduction of erosion, and pest incidences erosion (Dainese, Montecchiari, Sitzia, Sigura, & Marini, 2017). Information on the carbon storage in hedgerows is minimal. For example, no empirical research on carbon stocks for hedgerows in the UK has been published in scientific literature (Axe et al., 2017).
Notwithstanding that hedgerow carbon stocks are crucial to study because agroecosystems can provide various ecosystem services, including provisioning and regulating. Moreover, hedgerows can contribute largely to combating climate change by storing carbon. Carbon storage is a critical link in the global carbon cycle. After carbon dioxide is converted into organic matter by photosynthesis, a substantial pool of carbon is stored in woody biomass for a certain period of time before it is ultimately returned to the atmosphere through respiration and decomposition.
To elaborate, half of the global carbon sink is located in terrestrial ecosystems. Forest ecosystems account for 80% of the terrestrial carbon in biomass, which underlies their importance for carbons storage in the future (Karjalainen et al., 2003). This extensive knowledge of forest ecosystems enables a reliable biomass and carbons storage estimation, hence their contribution in reducing the amount of carbon dioxide in the atmosphere to reduce global climate change (Van Den Berge et al., 2021). However, woody biomass outside the forest, like hedgerows, is widely neglected in GHG reporting and in most European countries, it is only based on forest inventories (Eggleston, Buendia, Miwa, Ngara, & Tanabe, 2006). In addition, MacCarthy et al. (2015) report that the lack of quantitative information on hedgerows carbon stocks makes reporting their contribution to national greenhouse gas removals challenging.
Even if hedgerows are wanted in the national natural inventories, a suitable estimation mechanism is needed, but these were not available (Van Den Berge et al., 2021). Yet, the use of new and modern fine-scale measure techniques in biomass inventory has become increasingly efficient (Gollob, Ritter, & Nothdurft, 2020). Several studies have shown that Lidar, as an active remote sensing device, can accurately estimate biomass (Popescu, 2007a). Airborne laser scanning (ALS) has the advantage of being able to cover large regions (up to the regional level) (Gollob et al., 2020). Because of its ability to acquire ground returns over vegetated regions, airborne Lidar has been proven as the most optimal technology for obtaining accurate CHM over broad wooded regions 5. The biomass is then estimated using an empirical model combined with CHM data (Vazirabad & Karslioglu, 2011). However, ALS systems are often inadequate for extracting specific spatial information because the three-dimensional (3D) point clouds represent individual objects too sparsely (Gollob et al., 2020). Due to its capacity to show detailed spatial structures, mobile laser scanner (MLS) is used in forestry inventories to get a more detailed view of the vegetation (Gollob et al., 2020).
Research aim & questions
This study aims to quantify the carbon stored in the woody biomass of hedgerows and investigate the added value of Lidar data for estimating large-scale carbon storage in hedgerows. The study will focus on a trimmed hawthorn hedgerow as a case study. The hedgerow is situated on the border of the cow pasture of farm Hofstede Rhijnauwen. The study area is further referred to as Hofstede Rhijnauwen.
Hofstede Rhijnauwen is chosen because it is one of the few remaining hedgerow regions in the Netherlands. Hedgerows used to be a typical sight in the Dutch countryside (Roeleveld, ). According to a rough estimate based on topographic maps from 1900 to 1978, 70% of the hedgerows in the Netherlands have vanished (Roeleveld, ). Furthermore, the hedgerows are pruned into the same form twice a year, making calculating the overall volume simpler. Hofstede Rhijnauwen is situated in the municipality of Bunnik, the Netherlands (51° 04′ 25.2″ N, 5° 11′ 04.9″ E) (figure 1). The landscape is formed by the river course of the Kromme Rijn, which has resulted in a meandering pattern and surface-level differences. The characteristics of this area are the tight allotment and the bocage landscape (van Bergen, J. G. P., ).
To reach the aim, this research will focus on three questions: (i) what is the carbon storage per square meter of hedgerow (kg C m-2 hedge) and (ii) does Lidar data match with empirical data and (iii) what is the total carbon storage of the hedgerows of Hofstede Rhijnauwen? The first question will be answered by deriving a new method for quantifying woody biomass and gathering and integrating field data on hedgerow characteristics. The second question will be answered by analysing ALS height metrics and the inspection of HMLS spatial structure in RStudio to assess the accuracy of the Lidar data. For the third question, the ALS data is processed in ArcGIS combined with field data (i). The methods will be further explained in the methods and data section.
Figure 1: The study area near the city of Bunnik in the filled red dot and a photo of the sampled hedgerow on the right at the spot of the unfilled red dot (Base map provided by Esri, 2020).
Methods and data
Data
Field- and lab work data
In April 2021, twelve plots of hedgerow were established at the study site of Hofstede Rhijnauwen (figure 2). The nearest trunk was chosen for measurements every 10 meters, for a total of 120 meters. A trunk is divided into different orders (figure 3). The ordering is in accordance with Hack's ordering, originally meant for river streams. The main trunk of every sample is set to 1, and consequently, all its side branches receive order 2. Their side branches receive order 3, etcetera. The order of every trunk remains constant up to its initial link (Jasiewicz, ). The length, radius, number of side branches per order, number of trunks m-2 hedge, and the hedgerow geometry were measured; a total of 308 measurements.
In addition, 30 dm3 of hedgerow was destructively sampled by cutting the hedgerow. Three samples, one trunk and two order 2 side branches, were collected into plastic bags for temporary storage. The samples were then oven-dried at 105 degrees Celsius for 48 hours to estimate the dry weight. The volume was determined by the submersion method. To determine the specific gravity, dry mass is divided by volume.
Figure 2: Map of the sampled hedgerows, including the sample points as the yellow placemarks and the hedgerow segments in red (Base map provided by Google Earth Pro, 26 March 2007).
AHN & ZEB-REVO data
Lidar systems measure travel time between emitting a pulse and receiving a reflection of that pulse back to the sensor. The time measurement is then converted into a distance measurement to create a point cloud showing the reflected ground surface, including vegetation ((Popescu, 2007b)). Lidar instruments are categorized based on their platforms as (i) airborne- (ALS) and (ii) terrestrial- (TLS), and (iii) mobile laser scanning (MLS) (Haring, 2009).
For research question two, the accuracy of AHN3 data was analysed and further used in research question three to calculate the total carbons storage in the study area. AHN (algemeen Hoogtebestand Nederland) is an ALS based digital elevation map for the whole of the Netherlands (Kwaliteitsbeschrijving.). Additionally, ZEB-REVO data was used to assess the correctness of AHN3. The GeoSLAM ZEB-REVO RT (ZEB-REVO) 3D laser scanning system is based on MLS and captures data while an operator navigates the portable device across the sample plot, as opposed to ALS. As a result, ZEB-REVO data covers a smaller but more detailed area. For the metadata, see table 1.
Figure 3: Branch ordering in accordance to Hack ADDIN RW.CITE{{doc:60b3810 a8f08245c1e28f93e Jasiewicz,J [No Information]}}(Jasiewicz , ). Figure 3: Branch ordering in accordance to Hack (Jasiewicz, ).
Table 1: Overview of the metadata and its sources.
Subindex Platform Description Data type Publication date
Source
AHN3 Air-based Country width
height model of the Netherlands
Point cloud, Las
2014 POK
ZEB-REVO Ground-based Detailed 3d map of the study area (no green biomass)
Point cloud,
Las 2021 Thesis Supervisor: Dr. dr. D. A.C. Seijmonsbergen
Methods
Field- and lab work data
In order to answer the first research question, field data needed to be processed. For calculating the woody volume of the side branches (order 2, 3, 4) m-2 hedge, equation one was used. For calculating the woody volume of the trunk's m-2 hedge, equation two was used for the trunk. At last, equation three was used to calculate the total woody volume m-2 hedge.
𝑊𝑉𝑥= 𝜋 ∙ (𝑅𝑋2) ∙ 𝐿𝑋∙ 𝑁𝑥∙ 𝑁1𝑚2 Equation 1 𝑊𝑉1= 𝜋 ∙ (𝑅𝑥2) ∙ 𝐻𝐻 ∙ 𝑁1𝑚2 Equation 2 𝑊𝑉𝑡𝑜𝑡𝑎𝑎𝑙= 𝑊𝑉1+ 𝑊𝑉2+ 𝑊𝑉3+ 𝑊𝑉4 Equation 3 With: 𝐿𝑥: Mean length [m] 𝑅𝑥: Mean length [m] 𝑁𝑥: Mean radius [m]
𝑁1𝑚2: Mean number of trunks m-2 hedge [-/m2]
𝐻𝐻: Mean hedgerow height [m]
𝑊𝑉: Woody Volume m-2 hedge [m3/m2]
With (x) being the order; 2 or 3 or 4
Next, the woody volume m-2 hedge was translated to woody biomass m-2 hedge, with the fourth equation. This includes the specific dried weight, which has been determined in a laboratory.
𝑊𝐵 = 𝑊𝑉 ∙ 𝑆𝐺 Equation 4
With:
𝑊𝐵: Woody biomass m-2 hedge [kg/m2]
𝑆𝐺: Specific gravity [kg/m3]
At last, woody biomass m-2 hedge is translated to kg Carbon m-2 hedge. For this research, a value of 50% for the carbon content of dry wood and root is used (Matthews, 1993). Thomas and Martin (2012) investigated 253 woody species and concluded that wood carbon content varied between 43.4 and 55.6% for temperate species, including roots and branches.
AHN & ZEB-REVO data
A workflow was designed to answer the second and third research questions, including AHN and ZEB-REVO data (figure 4). This workflow involves a pre-process followed by an analysis to assess the accuracy of the AHN data, research question two, and an upscaling part to calculate the total carbon storage in the study area, research question three.
Fi g u re 4 : T h e wo rk fl o w fo r t h e p re -p ro c e s s , a n a ly s is a n d u p s c a lin g o f AHN a n d Z EB -REVO d a ta i n s o ft war e p ro g ra m s ; Ar c G IS, RStu d io (L id R p a c k a g e ) a n d Clo u d Com p a re (CC). W h e re b y L AS re fe rs to a L a s f o rm a t th a t s to re s t h e L id a r p o in t c lo u d d a ta .
Pre-process:
First, ZEB-REVO and AHN data needed to be adjusted. The dataset of AHN is available for the whole Netherlands and clipped to the study area in ArcGIS (step 1). Then both datasets are pre-processed in RStudio with the LidR package. For the scripts, see Appendix A. Next, AHN already included a ground classification, and therefore step 2 was skipped; otherwise, RStudio added more than two classes (ground & non-ground). Then, the ground points were normalized, and all points above two meters were excluded to remove overstory. The pre-processed data, named Las_new, could then be analysed.
Analysis:
Next, the AHN las files were clipped in CloudCompare to get three-point cloud segments of the hedgerow (figure 3). Metrics of the AHN were analysed since the third research question was based on this. Metrics are scalar summaries of point distributions and are calculated based on the point heights at the cell level of the point cloud (Roussel, Goodbody, & Tompalski, 2021). The metrics were calculated in RStudio with the LidR
Next, the ZEB-REVO and AHN las files are analysed based on visual inspection of the Canopy Height Model and 3D-plot to check the accuracy when hedgerows have nearby vegetation. A CHM is a raster layer representing the highest elevation of scanner returns, and a 3D plot is a rotatable viewer with points coloured by height coordinates (Roussel et al., 2021).
Upscaling:
At last, the AHN data, pre-processed, was loaded into ArcGIS, and 47 small hedgerow sections were created, roughly 28m2 each. With zonal statistics, the highest point in the point cloud per segment was selected. This height corresponds with the height of the hedgerow. These heights were then used in equation 2, combined with the results of the field data to get the total carbon storage in the hedgerows of Hofstede Rhijnauwen.
Results
Carbon storage m
-2hedge
In this part, the results of the first question are presented. The total woody volume m-2 hedge is 0.0325 m3/m2, with a mean hedgerow height of 1.38m and SD of the height of 10.25. In short, 2.35% of the total hedgerow volume is woody volume.
The mean specific weight of the sampled hawthorn is 717.56 kg m-3 with SD 70 (table 2). From this, it can be deduced that the woody biomass is 23.33 kg m-2 hedge, and the carbon content is 11.66 kg m-2 hedge, see table 3 for an overview of all values.
Table 2: Laboratory results on the specific weight of the three branch samples.
Dried weight (g) volume (cm3) density (kg/m3) 1. Trunk 86.02 122 705.08 2. Side branch 28.15 43 654.65 3. Side branch 14.67 18.5 792.97
Table 3: Results of the first research question; carbon storage m-2 hedge.
Variable Value
Woody volume m-2 hedge 0.0325 m3/m2
Wood density 717.56 kg/m3
Woody biomass m-2 hedge 23.33 kg/m2
Carbon content m-2 hedge 11.66kg/m2
Upscaling the results, 11.66 M kg of carbon is stored in 1 km2 of hedgerow, with a hedgerow height of 1.38 m. If the hedgerow height (HH) is different, the outcome of equation 2 changes and with that the carbon content.
Accuracy of Lidar data
AHN metrics
In this part, the results of the second question are presented. The derived metrics of the AHN data are accurate and match with the field data (table 4). For all measured heights, see Appendix B. The highest point (MAX) metric corresponds with the mean hedgerow height measured in the field. For segment 1; AHN gives a height of 1.312 meters and the field data 1.325 meters; this differs 1.3 cm, which is 1% of the actual hedgerow height. For segment 2; AHN gives a height of 1.27 meter and the field data 1.32 meter; this differs 5 cm, which is 4% of the actual hedgerow height. For segment 3; AHN gives a height of 1.501 meters and the field data 1.476 meters; this differs 2.5 cm, which is 2% of the actual hedgerow height. These results imply that the pre-process of ground classification and normalization of the AHN data gives the correct heights with only a small deviation above and below.
Table 4: Metrics of AHN data in RStudio and field data metrics.
Data Segment MAX HH (m)
AHN 1 1.312 AHN 2 1.27 AHN 3 1.501 Mean HH (m) Field 1 1.37 1.325 Field 2 1.4 1.32 Field 3 1.53 1.476
ZEB-REVO and AHN; Canopy Height Model and 3D-plot
The above three segments have little to no interference with nearby vegetation. The last used segment (not shown in figure 3) has a tree crown that hangs over the hedge. The height metric of the AHN is 1.98 meter; this is close to 2 meters, the height limit of the last file. In this case, the hedgerow height is not 1.98 meter, but the highest point is, which is part of the tree crown overhanging the hedgerow, see the black circle in figure 5 and 6. Excluding these points with the segment tool in CloudCompare, gives a hedgerow height of 1.92 meters. Thus, the height metrics were higher than desired due to the nearby foliage, implying that the assumed hedgerow height is inaccurate with nearby vegetation.
Total carbon Storage in the hedgerows of Hofstede Rhijnauwen
In this part, the results of the third question are presented. Subsequently, a map with heights of the 47 hedgerow segments is developed (figure 7). For the table with corresponding heights, see Appendix C. Referring back to the previous chapter, nearby vegetation may alter the metrics. Yet, using Lidar on a large scale, it may not be possible to check the accuracy manually. For these 47 segments, no adjacent segment's height metric differed significantly from the next. Therefore, no height metric is manually altered.
The total hedgerow area of Hofstede Rhijnauwen is 1330 m2 with a total carbon storage of 15516 kg. However, this is calculated with the average hedgerow height of 1.33 meter from the field data, which includes only a small section of the hedgerows. Including the variation of the heights for the total area of hedgerows gives a total carbon content of 15080 kg. This is 436 kg less carbon than the method with only one height. The smaller number results from a lower overall average hedgerow height; 1.29 meter instead of 1.33 meter.
Figure 6: CHM of AHN, whereby the highest point is visible in the black circle; overhanging tree crown. Colours represent the height (m).
Figure 5: 3D plot of ZEB-REVO, whereby the overhanging tree crown is visible in the black circle. Colours represent the height (m).
Figure 5: 3D-plot of ZEB-REVO, whereby the overhanging tree crown is visible in the black circle. Colours represent the height (m).
Fi g u re 7 :M a p o f h e d g e ro w s e g m e n ts wi th c o rre s p o n d in g AHN m a x im u m h e ig h t m e tri c s (m ). M AX re fe rs to t h e m a x im u m h e ig h t m e tri c . Col o u rs re p re s e n t h e ig h t. (B a s e m a p p ro v id e d b y E s ri , 2 0 2 0 ).
Discussion
In this section, the results of this research will be interpreted. In doing so, the three research questions that were presented in the introduction will be answered. This section will finalize by discussing possible improvements on the methodology and recommendations for further research.
Interpretation of the results
This research has further expanded the knowledge on carbon storage in hedgerow by quantifying the biomass of a trimmed hedgerow. The results point out that a trimmed hawthorn hedgerow with an mean height of 1.38 meters can store 11.66 kg of carbon m-2 hedge. 11.66 kg m-2 hedge, indicates that the reintroduction of hedgerows has significant potential to improve the delivery of regulating ecosystem services in the agricultural landscape, in addition to provisioning ecosystem services. Increasing the amount of carbon stores in agricultural lands has been mentioned by several initiatives and organizations as a crucial part of mitigating climate change (UNFCCC, 2015), and hedgerow reintroduction could be a part of the solution. This estimate of hedgerow carbon storage and the preceding derived method fill a knowledge gap on quantifying hedgerow biomass and carbon storage.
Woody biomass carbon stocks are higher than those found in a comparable study by Axe et al., (2017). Axe’s study is the only other empirical study to date, as he points out that empirical data on hedgerow carbon stocks in the UK is scarce. With a 40-year-old trimmed hedgerow, a height of 1.3m, and a mean diameter of 1.5cm, their data suggest a carbon stock of 3.22 kg/m2 hedge (8.13 stems m-2). The hedgerow samples in this study, in contrast, have a height of 1.38 meters and a mean diameter of 4.11 cm resulting in a carbon stock of 11.66 kg/m2 hedge (7.25 stems m-2). Furthermore, current models estimate hedgerow biomass to range from 0.5kg C/m2 to 4.5 C kg/m2, but these are no empirical studies (Falloon, Powlson, & Smith, 2004; Robertson, Marshall, Slingsby, & Newman, 2012). These models extrapolate values from average carbon stock values from other vegetation types. To highlight, Robertson et al., (2012) use hawthorn woodland understory stems of 1.3 m high and only diameters above 2.5 cm (0.2082 stems -2), giving a carbon stock of 4.5 kg m-2. Comparing Axe with Roberston; Roberston's hedgerow structure had a smaller area of woody growth comprised of fewer trunks but had a higher average diameter resulting in higher carbon stock. This suggests that the diameter of hedgerow stems is of large significance. As a result, despite the fact that the results of this study are greater than those of the comparison studies, it will most likely be slightly higher due to a larger mean trunk diameter (4.11 cm).
Next, the result of the first research question is partly generalizable, only for hedgerows with similar geometry. In more detail, 11.66 kg of carbon/m2 hedge results from a height of 1.38m. Suppose the height of a hedgerow is significantly higher than 1.38 m. In that case, the values of the variables used in the equations to calculate carbon content will change, resulting in a change in carbon content. For example, the amount of order 2 side branches on an order 1 trunk may increase when the order 1 trunk is higher. Thus, it is presumed that the result is only applicable to hedgerows of similar height. On the other hand, the developed method can be applied to hedges of various geometries as long as they are trimmed and flat on top.
The answer to the second research question is that ALS Lidar data is correct as long as adjacent vegetation does not interfere with the hedge causing unwanted points in the point cloud, altering the metrics. If more research is done on the accuracy for vegetated regions, ALS Lidar can be employed on a wide scale for hedgerow biomass and carbon storage inventories. This is significant since the usage of ALS Lidar remote sensing data in forestry has steadily increased over the last several decades (Sheridan, Popescu, Gatziolis, Morgan, & Ku, 2015) and Lidar data can be collected over larger areas with a reduced amount of effort compared to traditional field measurements (Clerici, Valbuena Calderón, & Posada, 2017).
The answer to the third research question is that the hedgerows in the study area store 15080 kg of carbon. In other words, already one farm with a hedgerow length of 1 km (width 1.33) can store approximately 15 Tons of carbon. Although the results only account for the area of Hofstede Rhijnauwen, it does show the potency of reintroducing hedgerows in other agricultural landscapes of the Netherlands. Reintroducing hedgerows on a larger scale would have more impact than reintroducing them on a small scale.
Methodological discussion
There are some points to improve this study. First of all, the wood density (717.56 kg/m3) is determined based on only three branches. However, in practice taking more samples was not justified. Besides, the wood density is lower than that on wood-database.com of 785 kg/m3 and timberpolis.co.uk of 800 kg/m3. The lab research could have been more extensive. For example, although one branch's diameter was more than double that of the other two, all three branches spent the same amount of time in the oven. This may have resulted in the thick branch being less dried and, as a result, having a higher mass, resulting in a lower specific weight. Also, the branches
were not entirely saturated with water during the submersion for volume determination; hence air bubbles remained in the pores. This means that there is still oxygen in the pores, implying that the force required to submerge the sample is larger. The rule states that the immersion force equals the (branch) volume. As a result, the volume of all branches is most likely exaggerated. Because the density of wood is mass divided by volume, the specific weights are likely overestimated and closer to the other values than previously thought.
Second, for this study, AHN is used and pre-processed. In this process errors could exist in determination of ground level (Bregt, Grus, van Beuningen, & van Meijeren, 2016). The default values of the implementation of ground classification are used in RStudio.
Third, during this research, several changes had to be made regarding the research questions and method. The limited time frame caused that the comparison between ZEB-REVO data and the AHN data were not addressed within this research. It has yet been studied, but eventually, it did not have an added value for answering the research questions.
Finally, the result of research question one (kg carbon m-2 hedge) has no standard deviation because it is a mean based on other means. However, a standard deviation may be derived by first calculating the carbon storage per plot (figure 3) and then taking the mean of the plots. However, the number of measurements per plot is too small, making the 12 separate carbon storages unreliable.
Outlook
Further research should focus on further developing the derived method for quantifying hedgerow biomass and carbon stocks. More quantitative data on hedgerow carbon stocks makes it easier to report their contribution to national GHG reductions and climate targets (Axe et al., 2017). Furthermore, the method could be improved to quantify the biomass of non-trimmed hedgerows in the future.
Also, before AHN Lidar can be accurately applied, the application of the metrics for hedgerow biomass predictions needs to be further investigated. For example, a tool like an outlier removal. Also of importance, in the first quarter of 2021, AHN4 will be released, which has a higher point density than AHN3 (Kwaliteitsbeschrijving.). It must be investigated whether a higher point density is beneficial for the height metrics.
At last, reintroducing hedgerows in the agricultural landscape asks for collaboration between multiple parties; scientists, government and farmers. Unfortunately, farmers are now discouraged from planting hedgerows on their land. Farmers do not receive a subsidy if they plant a hedgerow, but do receive it when they are removed and turned into cultivable land (Hakkenes, 2018). Therefore, a planting subsidy should be proposed for hedgerows in the agricultural land to encourage farmers to reintroduce hedgerows in their landscape.
Conclusion
This study aimed to quantify the carbon stored in the woody biomass of hedgerows and investigate the added value of Lidar data for estimating large-scale carbon storage in hedgerows. This is done on the hand of the three research questions: (i) what is the carbon storage per square meter of hedgerow and (ii) does LiDAR data match with empirical data, and (iii) what is the total carbon storage of the hedgerows of Hofstede Rhijnauwen?
The results show that reintroducing hedgerows is a potential solution for increasing carbon storage in the agricultural landscape. Because a trimmed hawthorn hedgerow with a height of 1.38 meter can store 11.66 kg of carbon /m-2 hedge. Whereby a new method for quantifying woody biomass is derived. Also, ALS Lidar data can be used on a large scale for hedgerow biomass inventory if the altered metrics due to surrounding vegetation are further investigated. At last, the combination of AHN data with field data results in a carbon storage of 15080 kg in the total 1 km long hedgerows of Hofstede Rhijnauwen.
A point to improve this study is to base the specific weight on more than three samples and perform lab work more precisely by completely saturating the samples and then adapting the drying period to the sample's volume. Hopefully, this case study encourages more extensive research on the quantification of hedgerow biomass and carbons stocks, including ALS Lidar data.
At last, the reintroduction of hedgerows in the agricultural lands asks for collaboration between scientists, the government and farmers. Scientists must effectively convey their expertise, and the government should encourage farmers to reintroduce hedges into the landscape by enacting a hedge subsidy policy.
References
References
Assessment, M. E. (2005). Ecosystems and human well-being Island press United States of America.
Axe, M. S., Grange, I. D., & Conway, J. S. (2017). Carbon storage in hedge biomass—A case study of actively managed hedges in England. Agriculture, Ecosystems & Environment, 250, 81-88.
Bregt, A. K., Grus, L., van Beuningen, T., & van Meijeren, H. (2016). No title. Wat Zijn De Effecten Van Een Open
Actueel Hoogtebestand Nederland (AHN)?,
Clerici, N., Valbuena Calderón, C. A., & Posada, J. M. (2017). Fusion of sentinel-1A and Sentinel-2A data for land cover mapping: A case study in the lower Magdalena region, Colombia. Journal of Maps, 13(2), 718-726.
Dainese, M., Montecchiari, S., Sitzia, T., Sigura, M., & Marini, L. (2017). High cover of hedgerows in the landscape supports multiple ecosystem services in M Mediterranean cereal fields. Journal of Applied
Ecology, 54(2), 380-388.
Eggleston, S., Buendia, L., Miwa, K., Ngara, T., & Tanabe, K. (2006). IPCC guidelines for national greenhouse gas inventories.
Falloon, P., Powlson, D., & Smith, P. (2004). Managing field margins for biodiversity and carbon sequestration: A great Britain case study. Soil use and Management, 20(2), 240-247.
FAO. (2021). Ecosystem services & biodiversity (ESB) . Retrieved from http://www.fao.org/ecosystem-services-biodiversity/background/regulating-services/en/
Gollob, C., Ritter, T., & Nothdurft, A. (2020). Forest inventory with long-range and high-speed personal laser scanning (PLS) and simultaneous localization and mapping (SLAM) technology. Remote Sensing, 12(9), 1509.
Hakkenes, E. (2018). Boeren willen subsidie voor het intact laten van bomen en bosjes op hun land | trouw. Retrieved from https://www.trouw.nl/duurzaamheidnatuur/boeren-willen-subsidie-voor-het-intact-laten-van-bomen-en-bosjes-op-hunland~bae47758
Haring, R. (2009). Airborne and terrestrial laser scanning. Retrieved from https://www.gim-international.com/content/article/airborne-and-terrestrial-laser-scanning
Holden, J., Grayson, R. P., Berdeni, D., Bird, S., Chapman, P. J., Edmondson, J. L., . . . Hunt, S. (2019). The role of hedgerows in soil functioning within agricultural landscapes. Agriculture, Ecosystems & Environment,
273, 1-12.
Jasiewicz, J.Stream ordering example . Retrieved from
https://grass.osgeo.org/grass78/manuals/addons/r.stream.order.html
Joos, F., Prentice, I. C., Sitch, S., Meyer, R., Hoss, G., Plattner, G., . . . Hasselmann, K. (2001). Global warming feedbacks on terrestrial carbon uptake under the intergovernmental panel on climate change (IPCC) emission scenarios. Global Biogeochemical Cycles, 15(4), 891-907.
Karjalainen, T., Pussinen, A., Liski, J., Nabuurs, G., Eggers, T., Lapveteläinen, T., & Kaipainen, T. (2003). Scenario analysis of the impacts of forest management and climate change on the European forest sector carbon budget. Forest Policy and Economics, 5(2), 141-155.
Kwaliteitsbeschrijving. Retrieved from https://www.ahn.nl/kwaliteitsbeschrijving
MacCarthy, J., Thomas, J., Choudrie, S., Passant, N., Thistlethwaite, G., Murrells, T., . . . Thomson, A. (2010). UK greenhouse gas inventory, 1990 to 2008. annual report for submission under the framework convention on climate change.
Matthews, G. (1993). The carbon content of trees.
Popescu, S. C. (2007a). Estimating biomass of individual pine trees using airborne Lidar. Biomass and Bioenergy,
31(9), 646-655.
Power, A. G. (2010). Ecosystem services and agriculture: Tradeoffs and synergies. Philosophical Transactions of
the Royal Society B: Biological Sciences, 365(1554), 2959-2971.
Robertson, H., Marshall, D., Slingsby, E., & Newman, G. (2012). Economic, biodiversity, resource protection and social values of orchards: A study of six orchards by the Herefordshire orchards community evaluation project. Natural England Commissioned Reports, (090)
Roussel, J., Goodbody, T. R. H. & Tompalski, P. (2021). The lidR package. Retrieved from https://jean-romain.github.io/lidRbook/index.html
Sheridan, R. D., Popescu, S. C., Gatziolis, D., Morgan, C. L., & Ku, N. (2015). Modeling forest aboveground biomass and volume using airborne LiDAR metrics and forest inventory and analysis data in the pacific northwest. Remote Sensing, 7(1), 229-255.
Thomas, S. C., & Martin, A. R. (2012). Carbon content of tree tissues: A synthesis. Forests, 3(2), 332-352.
UNFCCC. (2015). Join the 4/1000 initiative. soils for food security and climate.
van Bergen, J. G. P. Bunnik. Retrieved from https://mooisticht.nl/gemeenten/bunnik
Van Den Berge, S., Vangansbeke, P., Calders, K., Vanneste, T., Baeten, L., Verbeeck, H., . . . Verheyen, K. (2021). Biomass expansion factors for hedgerow-grown trees derived from terrestrial LiDAR. BioEnergy
Research, 14(2), 561-574.
Vazirabad, Y. F., & Karslioglu, M. O. (2011). LIDAR for biomass estimation. Biomass–Detection, production and
usage (pp. 3-26)
Acknowledgements
I would like to thank dhr. dr. D. A.C. Seijmonsbergen for his supervision and feedback throughout the project in addition to his help and skills with the ZEB-REVO data. Furthermore, I want to thank dhr. dr. K.F. Rijswijk for supporting me during the fieldwork and his inspiration on deriving a method for hedgerow biomass.
Appendices
Appendix A: RStudio code
clear clc # load packages --- getwd() library(lidR) library(raster) library(sp) library(rgdal) library(ggplot2) library(rgl) library(MBA) library(RCSF) # load data --- las<- readLAS("FILEPATH") # check data --- las_check(las) # basic plot --- plot(las, axis = TRUE, legend = TRUE)
# cross section 2d --- plot_crossection <- function(las,
p1 = c(min(las@data$X), mean(las@data$Y)), p2 = c(max(las@data$X), mean(las@data$Y)), width = 4, colour_by = NULL)
{
colour_by <- enquo(colour_by)
data_clip <- clip_transect(las, p1, p2, width)
p <- ggplot(data_clip@data, aes(X,Z)) + geom_point(size = 0.5) + coord_equal() + theme_minimal()
if (!is.null(colour_by))
p <- p + aes(color = !!colour_by) + labs(color = "")
return(p) }
plot_crossection(las, colour_by = factor(Classification))
# classify ground points ---
las <- classify_ground(las, algorithm = csf(sloop_smooth = FALSE, class_threshold = 0.5, cloth_resolution = 0.5, time_step = 0.65))
p1 = c(min(las@data$X), mean(las@data$Y)) p2 = c(max(las@data$X), mean(las@data$Y))
plot_crossection(las, p1 = p1, p2 = p2, colour_by = factor(Classification)) # height normalization ---
nlas <- normalize_height(las, knnidw())
plot_crossection(nlas, p1 = p1, p2 = p2, colour_by = factor(Classification)) # filter_poi below 2 m ---
las_bunnik_normalized_below2<- filter_poi(nlas, Z < 2)
writeLAS(las_bunnik_normalized_below2, file = "las_bunnik4_normalized_below2.las") # Canpoy Height Model ---
chm <- grid_canopy(las_bunnik_normalized_below2, res = 0.5, algorithm = p2r(subcircle = 0.3))
col <- height.colors(50) plot(chm, col = col)
# METRICS of nonground ---
las_bunnik_nonground<- filter_poi(las_bunnik_normalized_below2, Classification == 0) #CLASSIFICATION = 0 for ZEB-REVO, CLASSIFICATION = 1 for AHN
metrics <- cloud_metrics(las_bunnik_nonground, func = .stdmetrics_z) write.table(metrics, file = "metrics.csv")
Appendix B: Field data on hedgerow height
Table 5: Field data on hedgerow height divided into three segments.
Sample point
Hedgerow height (m)
Segment 1
1
1.36
2
1.37
3
1.37
4
1.20
Segment 2
5
1.22
6
1.34
7
1.40
Segment 3
8
1.40
9
1.53
10
1.50
11
1.48
12
1.47
Appendix C: Height metrics in ArcGIS
Table 6: Maximum height metric per hedgerow segment (m). MAX refers to the maximum height metric.
