The differences between the rainfall simulator and the mini disk infiltrometer, when measuring the infiltration rate in different soil types

(1)

________________________________________________________________________________

U

NIVERSITY OF

A

MSTERDAM

Facility of science

________________________________________________________

The differences between the rainfall simulator and the mini disk

infiltrometer, when measuring the infiltration rate in different

soil types

Bachelor Project Future Planet Studies

Thijs van Wieringen

E-mail: thijsvanwieringen@outlook.com

May 2021, Amsterdam

Institution: Biodiversity and Ecosystem Dynamics (IBED)

(2)

Institute of Biodiversity and Ecosystem Dynamics (IBED) University of Amsterdam

Future Planet Earth Science - Bsc project

Page 2 of 14

Abstract

When measuring infiltration rate in the field it is important that the equipment is easy to use, portable and affordable. Currently there are two often used methods which meet these criteria namely the rainfall simulator and the mini disk infiltrometer. However, little research has been done on the differences in results created by these two methods. Both methods measure the same value, but in a different way. The main goal of this research is to investigate if there is a significant difference in results between the rainfall simulator and the infiltrometer. Too add, this research investigated the amount of deviation and possible causes.

It became clear that there is a significant difference between the results from both methods. Results from the rainfall simulator were higher on average than those from the infiltrometer. This can be explained by the differences in size of both devices. The rainfall simulator being larger includes the presence of macropores and large cracks, while the infiltrometer only works on flat and even ground. Moreover, the rainfall simulator uses a drip plate, whereas the infiltrometer uses a sintered disk in combination with a suction regulator.

Although these methods show different results, they can indicate a trend in infiltration rates across different soils. Both methods having their pros and cons make them very compatible, and useful for future research. This data can be used when choosing which infiltration rate method is most suitable for the desired results. The researchers can consider the pros and cons of both methods, and weigh the differences in results against the costs, ease, and overall accuracy.

(3)

Inhoud

1.Introduction 4 2.Methods 5 2.1 Study area 5 2.1.1Groesbeek 5 2.1.2Hem 5 2.2 Infiltration methods 6

2.2.1Mini Disk Infiltrometer 6

2.2.2Rainfall Simulator 6

2.2.3 Soil physical visual analysis and Soil Moisture 7

2.2.4Statistical analysis 7

3.Results 8

3.1 Mini Disk Infiltrometer 8

3.2 Rainfall Simulator 8

3.3 Soil physical visual analysis and Soil Moisture 9 3.4 Mini Disk Infiltrometer vs Rainfall Simulator 9 3.4.1Statistical analysis Fout! Bladwijzer niet gedefinieerd.

4. Discussion 10

4.1 Mini Disk Infiltrometer 10

4.2 Rainfall Simulator 10

4.3 Soil physical visual analysis and Soil Moisture 11

4.4 Statistical analysis 11

5. Conclusion 12

(4)

Page 4 of 14

1. Introduction

Infiltration can be defined as the downwards entry of water into a soil or a sediment (Johnson, 1963 & Ferre & Warrick, 2005 & Haghnazari, Shahgholi, Feizi, 2015). When measuring the infiltration researchers are often interested in the infiltration rate, which is the amount of water that infiltrates the soil over time. Furthermore, the infiltration rate gives a general indication on the state of the soil and how it will deal with heavy rainfall or a storm.

This subfield of hydrology studies has gained attention in the last decades due to its importance in irrigation, contaminant transport, groundwater recharge and overland flow (Ferre & Warrick, 2005 & Assouline, 2013). Moreover, it is predicted that the need for knowledge about infiltration increases, with the increasing variability in extreme weather events due to climate change (IPCC, 2012). Therefore, it is crucial to gather more quantitative data on the infiltration capacity of soils to explain current hydrologic processes and use it to predict future systems undergoing hydrological change (Assouline, 2013).

Infiltration rate is usually determined from field data because complex soil structures and differences in soil properties make it difficult to estimate infiltration rates by using an equation or comparing it to previous measurements (Krimgold & Beenhouwer, 1954 & Johnson, 1963). The field data can be obtained using a variety of methods, all having their pros and cons. Two frequently used methods are sprinkling the soil with a rainfall simulator or using an infiltrometer. Both methods are easy to setup, fast and cost efficient, which makes them widely used in this field of research. However, there are also differences between the two methods.

The infiltrometer comes in various shapes and sizes, which makes it suitable for most fieldwork researches. It also creates accurate saturated permeability readings, on different soil profiles due to its suction regulator. However, according to Johnson (1963), the most relevant measurements are produced with a large ring infiltrometer, which are expensive and bigger compared to a mini disk infiltrometer with a surface area of approximately 5 centimetres. Although large rings are less practical, they will average the infiltration rate and are therefore less vulnerable for errors caused by large cracks, rocks or three roots.

The rainfall simulator is a device which measures the infiltration rate by recreating natural rainfall as good as possible on a soil sample taken from the field (Ngasoh, 2020). This method is often used when the interest lays primarily on determining the runoff, the counterpart of the infiltration, of the soil. In addition to the infiltrometer, this device can also estimate erosion rates of the surface material. Moreover, the rainfall simulator includes the kinetic energy and the random distribution of raindrops.

The correct method to use depends on the desired results, budget, and the field conditions. Although both methods are proven to be accurate in measuring infiltration rates, they use a different approach to obtain this. That is why in this research I will investigate whether there is a difference in results from both methods, and if so, what causes these differences.

First, I will use a statistical test to determine if there is a significant difference in measurements from the rainfall simulator and the infiltrometer. Next, I will determine which explanatory variable is most responsible for the deviation using statistics. Furthermore, the soil moisture content will be measured, which can be used to explain possible fluctuations in results due to the difference in weather conditions during the fieldwork days. At last I will combine this data with literature to possibly explain my results.

Both methods will be tested on three soil types containing primarily clay, sand, or loam, which can be found in the two study areas in Hem and Groesbeek. To make the results of this research useful, I will focus on the starting researchers with a small budget. I have therefore chosen to test a mini disk infiltrometer, and a simple portable rainfall simulator. Both methods are easy to use, cheap and they are suitable for all soil types.

(5)

2. Methods

2.1 Study area

For this research three soil types were chosen namely: clay, sand or loam. This will provide for a range of results, from compact clay to loose sand. Measurements from these three soil types will be collected at two locations. The most compact soil will be a clayey soil and can be found in Hem, Noord-Holland. The other two soil textures (Loess and sand) can be found on the moraine; Mokerheide, close to Groesbeek.

2.1.1 Groesbeek

The study area Groesbeek is located on the edge of the German and Dutch border in the province of Gelderland (figure 2). The area is hilly and has a height difference of approximate 50 meter. In the before last glacial period (130,000-110,000 years ago) a moraine has formed here containing coarse sand and gravels. Later, during the tundra circumstances, polar winds deposited loess on top of these hills (Stiekema, 2012).

In this area two soil types are predominant. The first and foremost soil is a Loamy soil comprised of 50-85% loam (Stolte et al., 1999). Loess is an aeolian deposit of silt, which means a grain size of <63 µm, > 2 µm (Hiscott, Hall & Primer, 1997). The type of soil found here is perhaps more well known as periglacial Loess. Furthermore, sandy soils can also be found either on the surface or underneath the Loess layer. Sand has a much greater texture with grain size particle being >63 µm.

Soils containing primarily loess hold water well, contain high amounts of organic matter and are therefore suitable for agriculture (Cat, 2001). However, due to the small grain size of loss water tends to infiltrate slowly in these soils. As a result, the area experiences overland flow and mudflows during heavy rainfall.

2.1.2 Hem

The second study area is Hem, a site where marine clay is the dominant soil type. Due to the high compactness of

this soil type, permeability is expected to be the lowest of three (Van Maanen et al., 2020). This makes for an interesting contrast with the sand and loess in Groesbeek. The measurements were taken on a

farm where cauliflower is cultivated. This farm often collaborates with the University of Amsterdam, and many researches have been conducted here in the past 5 years. When the measurements for this research were taken almost no rain had fallen in the last 2 weeks. The soil was very dry, and the clay was cracked in most places creating large macro pores.

Figure 1:Spatial reference for both study areas. Represented in blue

is marine clay (ISRIC World Soil Information/Naturalis).

Figure 2: Marked sample locations from Groesbeek and Hem. The colour of the marker represents the soil type. Respectively yellow, red, and blue is sand loess and clay (Google Earth, 2021).

(6)

Page 6 of 14

2.2 Infiltration methods

2.2.1 Mini Disk Infiltrometer

The infiltrometer measures the unsaturated hydraulic conductivity, by using the water level difference in the cylinder over time. Special about this meter is that it measures the hydraulic conductivity at different applied tensions. This can be advantageous over other methods because soils are often non uniform and have large macropores. These pores generally fill with air, hereby influencing the results (Rose 1966; Brady and Weil, 1999). Furthermore, the infiltrometer has an adjustable suction (table 1). A small negative pressure or suction prevents water from entering macropores, which ensures an accurate reading of the unsaturated hydraulic conductivity (Infiltrometer User’s Manual, Decagon, USA 2007). Using the instructions of Decagon Devices, the data is gathered for the different soil types. After data collection, the infiltration can be calculated using the provided macro (appendix 1.b).

Groesbeek1: Sandy soil Groesbeek2: Loess soil Hem: Clayey soil

Suction Rate (cm) 6 2 0.5

Time Interval (s) 2-5 30 180-360

Samples/location 1 4 1

Repetitions 5 20 (4x5) 5

2.2.2 Rainfall Simulator

In this research the rainfall simulator designed by Wageningen University will be used. The simulator is placed on a platform 0.5m off the ground, with a jerry can on top providing water. It uses a drip plate to simulate rain and an adjustable valve in combination with a perforated cork to set the intensity. A metal ring is placed in the ground to demarcate the area under investigation and to prevent lateral movement (Figure 3). For this research a low rain intensity will be used, since this is most realistic for rainfall in the Netherlands. Each test will follow the same procedure (Figure 4), with the same settings (Figure 5). Finally, the runoff data will be combined with the rain intensity to calculate the infiltration (Figure 6).

Figure 3: Rainfall Simulator (Cammeraat, 2021)

Procedure Rainfall Simulator

1. Place the ring on a slightly sloped surface and cover the insight with plastic for possible spillage when placing the drip plate on top.

2. Fill the jerrycan with water and set the intensity with one of two corks (81 mm/hr, 454 mm/hr).

3. Remove the plastic and start the simulation.

4. Note the time to runoff and place a container to collect the runoff under the outlet.

5. Renew the bottle every 2 minutes until the runoff flattens for 5 bottles. If the runoff is constant, stop the experiment. 6. Perform the experiment 5x per location to get accurate

results.

Procedure Rainfall Simulator

7. Place the ring on a slightly sloped surface and cover the insight with plastic for possible spillage when placing the drip plate on top.

8. Fill the jerrycan with water and set the intensity with one of two corks (81 mm/hr, 454 mm/hr).

9. Remove the plastic and start the simulation.

10. Note the time to runoff and place a container to collect the runoff under the outlet.

11. Renew the bottle every 2 minutes until the runoff flattens for 5 bottles. If the runoff is constant, stop the experiment. 12. Perform the experiment 5x per location to get accurate

results.

Figure 4: Summary of procedure used to measure infiltration with the Rainfall Simulator.

Rainfall intensity (low) 81 mm/hr

Rainfall intensity (high) 454 mm/hr

Surface area ring 7.07 dm2

Fall distance raindrops 50 cm

Volume jerrycan 9.0 L

Estimated duration one experiment 40 min

Rainfall intensity (low) 81 mm/hr

Rainfall intensity (high) 454 mm/hr

Surface area ring 7.07 dm2

Fall distance raindrops 50 cm

Volume jerrycan 9.0 L

Estimated duration one experiment 40 min

Figure 5: Specifications Rainfall Simulator designed by Wageningen University.

Table 1: Mini disk infiltrometer adjustment, suction rate per soil texture. The amount of locations sampled and repetitions per soil sample area.

Groesbeek1: Sandy soil Groesbeek2: Loess soil Hem: Clayey soil

Suction Rate (cm) 6 2 0.5

Time Interval (s) 2-5 30 180-360

Samples/location 1 4 1

Repetitions 5 20 (4x5) 5

Table 1: Mini disk infiltrometer adjustment, suction rate per soil texture. The amount of locations sampled and repetitions per soil sample area.

(7)

2.2.3 Soil physical visual analysis and Soil Moisture

For each sampling location a soil profile will be taken of 1,20 m in depth, where if needed an extension part is present to increase the length to 2.20 m. To add, the World Soil Reference Base (WRB) will be used to classify the soil profile (figure 7&8). This data can later be used in combination with the infiltration results to explain the hydrological behaviour of the soil.

Additionally, the moisture content will be measured with the ThetaProbe soil moisture sensor, both before and after water a rainfall simulation. This sensor measures the volumetric soil moisture content by responding to changes in the apparent dielectric constant. Then, the moisture content can be defined as the ratio between the volume of water to volume of dirt. Finally, linear interpolation will be used to determine the moisture content for the different measurement (Appendix 1.c).

It is interesting to measure this because an unsaturated soil has a higher sorptivity. When the soil is dry, water can infiltrate more easily until the saturated state has been reached. Once Saturated hydraulic conductivity has been reached, water starts to only infiltrate through gravitational forces and the suction tensions no longer plays a part (Cammeraat 2021).

2.2.4 Statistical analysis

For this research a two-way anova test will be performed in combination with a multiple regression test. The data has been tested for normality (Shapiro-Wilk test) and Heteroscedacity (Bartletts test), for which they both passed. The two-way anova test shows if there is a significant difference between the two groups by looking at the means, explained by the treatment (Rainfall Simulator or Infiltrometer) and the Soil types (Sand, Loam and Clay). Next, a Tukey test is performed to see which groups are most significant. Lastly, multiple regression will be used to test how much variance is determined by the explanatory variables (Appendix 1.d).

Infiltration (mm/h) = Rainfall (mm/h) – Runoff (mm/h)

Runoff (ml/min) → Runoff (L/min) → Runoff (dm3_/min)→𝑅𝑢𝑛𝑜𝑓𝑓 (𝑑𝑚3/𝑚𝑖𝑛)

𝑆𝑢𝑟𝑓𝑎𝑐𝑒 𝐴𝑟𝑒𝑎 (𝑑𝑚2)→ Runoff (dm/min)→ Runoff (mm/hr)

Runoff (ml/min) → Runoff (L/min) → Runoff (dm3_/min)→𝑅𝑢𝑛𝑜𝑓𝑓 (𝑑𝑚3/𝑚𝑖𝑛)

𝑆𝑢𝑟𝑓𝑎𝑐𝑒 𝐴𝑟𝑒𝑎 (𝑑𝑚2)→ Runoff (dm/min)→ Runoff (mm/hr)

Rainfall (ml/min)→ Rainfall (L/min)→ Rainfall (dm3_/min)→𝑅𝑎𝑖𝑛𝑓𝑎𝑙𝑙 (𝑑𝑚3_{/𝑚𝑖𝑛)}

𝑆𝑢𝑟𝑓𝑎𝑐𝑒 𝐴𝑟𝑒𝑎 (𝑑𝑚2₎ → Rainfall (dm/min) → Rainfall (mm/hr)

Rainfall (ml/min)→ Rainfall (L/min)→ Rainfall (dm3_/min)→𝑅𝑎𝑖𝑛𝑓𝑎𝑙𝑙 (𝑑𝑚3_{/𝑚𝑖𝑛)}

𝑆𝑢𝑟𝑓𝑎𝑐𝑒 𝐴𝑟𝑒𝑎 (𝑑𝑚2₎ → Rainfall (dm/min) → Rainfall (mm/hr)

Figure 6: Calculating Infiltration for the Rainfall Simulator from Rainfall intensity (mm/hr) and runoff (mm/hr) (Cammeraat, 2005).

Figure 7: Soil texture triangle to determine the texture of the soil (NZ

Soils, 2021)

Figure 8: The texture of the soil can then be grouped into one of four main categories. Texture group triangle (NZ

(8)

Page 8 of 14

3. Results

The result section is divided into four sections. The first two sections give the main results of the infiltrometer and the rainfall simulator. The results from the soil profiles and the soil moisture are given in section three. In the last section both methods are compared, and differences between the two are calculated. In this section differences are also analysed in a statistical way to find out if the two methods differ significantly.

3.1 Mini Disk Infiltrometer

The infiltrometer has been used for five repetitions on six locations, together forming thirty measurements. The sandy soil has the highest average inflation rate of 39.94 mm/hr, whereas the lowest mean value is measured at location 2, where loam is the dominant soil type. Although the lowest mean infiltration has been found in loamy soil, the absolute lowest 3 values have been measured on the clay location. However, the standard deviation on this location is high (9.43 mm/hr), which makes that this soil doesn’t have the lowest mean infiltration.

3.2 Rainfall Simulator

For the rainfall simulator 21 samples were taken on a total of 6 locations. The samples from Hem are excluded in this section due to dry weather conditions. The soil was too dry, and runoff didn’t occur after 45 minutes of testing with both intensities. The saturated hydraulic conductivity hasn’t been reached within this time. On the y-axis the infiltration rate has been set, with a maximum of 76.21 mm/hr and a minimum of 43.06 mm/hr, both on a loamy soil. Two tests were performed for location 2-6, whereas location 1 has one more with three repetitions.

Figure 9: Infiltration rate (mm/hr) measured with the Mini Disk infiltrometer (created in Microsoft Office 365 Excel).

Figure 9: Infiltration rate (mm/hr) measured with the Mini Disk infiltrometer.

Figure 10: Infiltration rate (mm/hr) measured with the Rainfall Simulator. Note that the clay soil has been set to a value > 81.19, so the bar isn’t in

perspective (created in Microsoft Office 365 Excel).

Figure 10: Infiltration rate (mm/hr) measured with the Rainfall Simulator. Note that the clay soil has been set to a value > 81.19, so the bar isn’t in

(9)

3.3 Soil physical visual analysis and Soil Moisture

In table 2 the results of the soil analysis and the soil moisture content are displayed. Three different texture classifications have been found, with underneath the composition of Sand Loam and Clay. Because the soil groups are broad in term, different compositions can still translate to the same terminology (Appendix 1.e).

Moreover, the presence (Y) or absence (N) of a dry, cemented, and crusty top layer as well as macropores or deep cracks have been described. An impermeable layer blocks water flow into the subsoil, which lowers the infiltration rate. Conversely, the infiltration rate can be higher than expected when macropores or large cracks have formed in the topsoil.

The soil moisture content before the simulation ranges from only 9% to 30% water to volume of soil. This wide range indicates that some soils were dry at the start, while others already contained some water. Values obtained after running the simulation lay closer together ranging from 0.28% to 0.41%. The highest increase in moisture content has been found on the sandy soil of location 5, while the lowest was obtained on the loamy soil in location 2.

3.4 Mini Disk Infiltrometer vs Rainfall Simulator

When comparing figure 9 and 10, it becomes clear that there is a difference in infiltration between the infiltrometer and the rainfall simulator. Across all location the rainfall simulator shows a higher mean infiltration rate than the infiltrometer. The largest deviation has been measured in location 3, where the saturated conductivity differs by 63.12 mm/hr.

However, one cannot already conclude that the two methods are significantly different from each other. Therefore, a two-way anova has been computed where the treatment and soil type have been used as explanatory variables.

The anova analysis yielded a main effect for ‘method used’, F(2,23) = 54.35, p = 1.69e-07, such that the average infiltration rate is higher for the rainfall simulator (M = 56.09 mm/hr, SD = 15.29) than for the infiltrometer (M = 25.22, SD = 11.83). The main effect of soil type was less significant with a F-ratio of F (1,23) = 6.84, p = 0.00468 (Appendix 1.d).

Moreover, method used significantly predicted infiltration rates, b=32.16, t (23) = 7.37, p<.01. From the soil types only, the loamy sand significantly predicted infiltration rates, b=25.70, t (23) = 7.552, p<.01, whereas the other two soils were not significant. Method used explained a significant proportion of variance in infiltration rates, R2_{= .75, F (3,23) = 22.68, p<0.01.}

Lastly, the post hoc tukey test showed that the rainfall simulator and the infiltrometer significantly differed (p = 2e-07), while from the soil types only between sandy loam and sand was significant (p = 0.0034).

Table 3: A brief overview of the results from the soil analysis and the soil moisture content. The composition has been given in percentages, where 55-45-0 stands for a sandy loam soil containing 55% sand, 45% loam and 0% clay. The soil moisture has been given in percentages from total volume of the

(10)

Page 10 of 14

4. Discussion

4.1 Mini Disk Infiltrometer

When using the infiltrometer to measure infiltration rates across different soil types, a few things stood out. Most prominent was the size, ease of use and overall simplicity of this measuring device. Although its simplicity, this method of measuring infiltration has managed to generate some interesting results.

The data first had to be calculated and converted into the desired form (mm/hr). A macro provided by the founder of this device was used. In here the suction rate and soil type had to be filled in manually. This was needed for altering the results created by the formula. It was after this step that the results started to make more sense. As can be seen in figure 9, a difference can already be seen between different soil types, which hasn’t been noticed by looking at the data at first. The highest infiltration rate can be linked to the sandy soil, which according to Meek, Rechel, Carter, DeTar, & Urie, (1992), is to be expected. Due to the larger grain size, water can infiltrate more easily into these soils. On the other hand, infiltration rate was lower in loamy soils, which can be related to smaller grain size (Meek et al., 1992).

Although the infiltrometer has been accurate in results, and showed the general differences between soil types, it also lacked the ability to repeatedly create the same value on the soil. On half of the locations there was a high standard deviation, which was created by outliers. Location 4 for example has a difference of over 30 mm/hr between the lowest and highest measurement, which is to large in this context. The manual proscribed the use of five repetitions, however after these results, it might have been better to increase this by a few.

Moreover, the size of the disk is of importance for this deviation (Johnson, 1963), as well as the surface on which it has been placed on (Alinea 4.3). The small surface of the disk makes it so that the device cannot be placed on a surface with small cracks, macropores or vegetation. This results in a selective behaviour when placing the device for measuring, and therefore creates only results of small areas with a flat surface. Outliers have probably been created by placing the infiltrometer on a non-flat surface, which makes it drain to quick.

Lastly, the suction rate had to be adjusted according to the soil (Infiltrometer User’s Manual, Decagon, USA 2007). Therefore, the soil type had to be determined first. Often the soil type was questionable or not perfectly fitting one terminology, but still one had to be chosen to set the suction. A difference in suction rate generates a vastly different result, which could lead to error when making conclusions. Admittedly it might have been better to put the suction rate not to either one, but in the middle to compensate for the uncertainty.

4.2 Rainfall Simulator

The rainfall simulator is larger, takes more time to setup and is therefore less desirable for field trips. Still, some interesting results were created using this method. An important result was that there was a relatively large difference between the repetitions on each location.

First, this could be due to a mismatch of the drip plate and the ring. When setting up this method, one must place the drip plate exactly above the ring, to obtain maximum infiltration. However, the ring and drip plate are not the same shape and size. This results in a potential loss of rain intensity which falls outside the ring (and doesn’t create runoff this way). When calculating the infiltration with this setup, a larger rain intensity is being used than was in fact applied.

Additionally, puddles were created inside the ring, which lowered the runoff value. The puddles were caused by uneven ground or large plant matter, which blocked a path to the runoff collection container. Sometimes the ‘dam’ (a stick blocking the runoff stream) would burst, and runoff would increase drastically, whereas sometimes the value was very low because puddles were getting filled. As a result, the runoff values were highly fluctuating during the run. This in term has led to differences in infiltration values.

Moreover, something occurred to me while performing the runs of the rainfall simulator. When running it for longer periods (close to one hour), the rain intensity seems to drop slightly when time progresses. A possible explanation is that the water pressure of the holding tank decreases when more water leaves the tank overtime. In the beginning, the pressure is high, and the rain intensity is also high, but whenever the water level was low, the rain intensity seems to slow down as well. Nevertheless, this was only an observation which hasn’t been further investigated afterwards.

Also, the presence of deep cracks and macropores can lead to a higher infiltration. According to Germann, Edwards & Owens (1984), infiltration of rainwater is higher when macropores are present in the soil. It is stated that the water infiltrates faster, and deeper into the lower sublayers which often can transport the water more quickly.

Lastly, something to be noted is that for this research a high and low intensity was used. However, while measuring it became clear that the high intensity was not relevant for this study. To improve results for future research even more only the low intensity of 81.19 should be used, since higher rainfall isn’t applicable for this

(11)

climate. Unfortunately, there was no time left to do more low intensity measurements for this research since one simulation takes over one hour to perform.

4.3 Soil physical visual analysis and Soil Moisture

Linking the soil structure and soil moisture to the infiltration rates also creates interesting relationships. The soil structure can in some cases determine if a soil lets trough more, or less water than to be expected. Moreover, the soil moisture content at the beginning of the experiment will influence the sorptivity of this soil. When the soil is dry, sorptivity increases and the ground will take more rainwater until deficiencies are replenished (Green, Ahuja & Chong, 1986).

As mentioned before, there were some locations that had a dry, cemented, and crusty top layer. When combining the infiltrometer data (figure 9) and the presence of a crusty top layer (table 3), it can be noted that high standard deviation always co-exists with this top layer. In an article from Eckert, Wood, Blackburn & Peterson (1979) on infiltration rate change due to off-road vehicle activity, they found that infiltration can be increased significantly when cars break up the top layer, hence allowing more water to infiltrate. Comparing this to the results of this thesis, one can also see that infiltration rates are influenced by this top layer.

4.4 Statistical analysis

From the anova test it becomes clear that the results of the rainfall simulator are significantly higher than those of the infiltrometer. According to multiple studies this is to be expected, because the infiltrometer only measures a small area, this way excluding the complexity of soil structures (Beven and Germann, 1982; Nieber and Sidle, 2010). The rainfall simulator uses a metal ring that is large enough to include macropores, differences in organic matter content, presence of vegetation, and variability in compaction. The device will therefor average out on most outliers of the soil structure and its complexness. This can also be seen in the results because standard deviation was lower in the results from the rainfall simulator than from the infiltrometer.

(12)

Page 12 of 14

5. Conclusion

In this research a significant difference has been found in the results created from the rainfall simulator and the mini disk infiltrometer. The results from the rainfall simulator were higher in each location compared to those of the infiltrometer. Looking at future research, people should be aware of the variability in results and potential error created by these two methods. Because the exact infiltration rate of the sample area is unknown, it is unclear which method is most accurate. Further research is needed to find out which device creates most reliable results across common soil types.

Although it is uncertain which method is most accurate, both devices showed the general trend and differences across soil types, which makes both methods very competitive. The infiltrometer is easier to use, takes less time to generate results, is water efficient, affordable and the most portable of all infiltration measurement devices. But in return, it only measures a small surface area which makes it vulnerable for creating outliers. The rainfall simulator is bigger, and overall requires more work. But it includes the randomness and kinetic energy of raindrops, and it samples a much larger area.

Too conclude, the device to use depends on the situation and the goal of the experiment. The rainfall simulator is more appropriate when infiltration is measured, but there is also a side interest in runoff. The infiltrometer is much lighter and water is not an issue. This device is likely to be used to quickly get an indication of the infiltration while being on a fieldwork, or somewhere where access to water is limited.

(13)

6. Reference list

Assouline, S. (2013). Infiltration into soils: Conceptual approaches and solutions. Water Resources Research, 49(4), 1755-1772.

Bhave, S., & Sreeja, P. (2013). Influence of initial soil condition on infiltration characteristics determined using a disk infiltrometer. ISH Journal of Hydraulic Engineering, 19(3), 291-296.

Cammeraat, L.H., personal communication, March 18, 2021 Cammeraat, L.H., (2005). Recondes Field Protocol, vs. 1.0

Catt, J. A. (2001). The agricultural importance of loess. Earth-Science Reviews, 54(1-3), 213-229. Decagon Devices, Inc. Mini Disk Infiltrometer, 2016. Version: September 2, 2016 — 13:13:56.

Delta-T Devices Ltd. ThetaProbe SOIL MOISTURE SENSOR. TYPE ML2x USER MANUAL ML2x-UM-1.21, May 1999. Eckert, R. E., Wood, M. K., Blackburn, W. H., & Peterson, F. F. (1979). Impacts of off-road vehicles on infiltration and

sediment production of two desert soils.

Ferre, T. P. A., & Warrick, A. W. (2005). Infiltration, Editor (s): Daniel Hillel, Encyclopedia of Soils in the Environment. Germann, P. F., Edwards, W. M., & Owens, L. B. (1984). Profiles of bromide and increased soil moisture after infiltration

into soils with macropores. Soil Science Society of America Journal, 48(2), 237-244.

Google Maps, 2021. UvA: University of Amsterdam, 1:20000. Google Maps [online] Available through: https://www.google.nl/maps/@52.6478688,5.2008812,417m/data=!3m1!1e3 [Accessed 26 April 2021].

Green, R. E., Ahuja, L. R., & Chong, S. K. (1986). Hydraulic conductivity, diffusivity, and sorptivity of unsaturated soils: Field methods. Methods of Soil Analysis: Part 1 Physical and Mineralogical Methods, 5, 771-798.

Haghnazari, F., Shahgholi, H., & Feizi, M. (2015). Factors affecting the infiltration of agricultural soils. International Journal

of Agronomy and Agricultural Research, 6(5), 21-35.

Hiscott, R. N., Hall, F. R., & Pirmez, C. (1997). Turbidity-current overspill from the Amazon Channel: texture of the silt/sand load, paleoflow from anisotropy of magnetic susceptibility, and implications for flow processes. In Proceedings-Ocean Drilling Program Scientific Results (pp. 53-78). NATIONAL SCIENCE FOUNDATION.

IPCC, 2012 – Field, C.B., V. Barros, T.F. Stocker, D. Qin, D.J. Dokken, K.L. Ebi, M.D. Mastrandrea, K.J. Mach, G.-k. Plattner, S.K. Allen, M. Tignor, and P.M. Midgley (Eds.) Available from Cambridge University Press, The Edinburgh Building, Shaftesbury Road, Cambridge CB2 8RU ENGLAND, 582 pp. Available from June 2012

ISRIC World Soil Information/Naturalis. Zeekleibodem voorkomen in Nederland. Retrieved from https://www.geologievannederland.nl/ondergrond/bodems/zeekleibodem-zeekleilandschap

Johnson, A. I. (1963). A field method for measurement of infiltration. Washington: US Government Printing Office. Kamphorst, A. (1987). A small rainfall simulator for the determination of soil erodibility. NJAS wageningen journal of life

sciences, 35(3), 407-415.

Krimgold, D. B., & Beenhouwer, O. (1954). Estimating infiltration. Agr. Eng, 35(10), 719-725.

Meek, B. D., Rechel, E. R., Carter, L. M., DeTar, W. R., & Urie, A. L. (1992). Infiltration rate of a sandy loam soil: effects of traffic, tillage, and plant roots. Soil Science Society of America Journal, 56(3), 908-913.

Ngasoh, F. G., Mbajiorgu, C. C., Kamai, M. B., & Okoro, G. O. (2020). A Revisit of Rainfall Simulator as a Potential Tool for Hydrological Research. In Agrometeorology. IntechOpen.

Robinson, A. R., & Rohwer, C. (1957). Measurement of canal seepage (Doctoral dissertation, Colorado State University. Libraries).

Stiekema, M. (2012). Archeologisch bureauonderzoek Plangebied De Mortel te Mook in de gemeente Mook en Middelaar. Econsultancy.

Stolte, J., C.J. Ritsema, T. Li. Invloed van verschillende landinrichtingsscenario 's op de bodemen waterafvoer in het zuidelijke deel van de ruilverkaveling Groesbeek, 1999. Wageningen, DLOStaring Centrum. Rapport 644. 54 blz. 24 fig.; 8 tab.; 10 ref.

Van Maanen, R., Morrien, E., & Cammeraat, E. (03-02-2020). Een vierjarig onderzoek naar het effect van niet onderwerken van strorijke stalmest in de bloemkoolteelt van Noord-Holland. Amsterdam Green Campus, duurzame innovatie in de regio.

Zhang, R. (1998). Infiltration models for the disk infiltrometer. Oceanographic Literature Review, 6(45), 941. Figures

Figure 4: Soil texture triangle, retrieved from: http://www.nzsoils.org.nz/Topic-Describing_Soils/Using_the_Texture_Triangle/ (15-03-2021)

Figure 4: Soil texture triangle, retrieved from: https://www.researchgate.net/figure/Canadian-soil-texture-triangle-that-uses-only-two-axes-to-determine-soil-texture-from_fig3_50917086 (15-03-2021)

(14)

Page 14 of 14

Appendix 1.a

Data Fieldwork Hem & Groesbeek. (25 Maart 2021 ~ 31 Maart 2021)

By clicking on the link, access is granted to a public folder with the fieldwork data used in this research. Also, the Excel files are included which have been used for calculation, creating figures and statistics. Most subfolders have Read.me files, used to explain what is in the files, what they have been used for and or what they mean.

https://drive.google.com/file/d/11S-HVp-KbC0QdjRbqmhVmbSmCV-tGx-3/view?usp=sharing

1.b

Pathway Macro Infiltration Calculation Mini-Disk Infiltrometer Appendix 1a 

(InfiltrationData_Groesbeek_Hem_28_05_2021\Combined_Data\Extra\Macro_Mini_Disk_Infiltrometer)

1.c

Pathway Calculation Soil Moisture Content

Appendix 1a ➔ InfiltrationData_Groesbeek_Hem_28_05_2021\Combined_Data\Summary\Thijs When opened ➔ Tab 4 ‘Soil Moisture Calculation’

1.d

Pathway Statistics Anova Test, Multiple Regression, Tukey, Shapiro-Wilk test and Bartletts test Appendix 1a ➔ InfiltrationData_Groesbeek_Hem_28_05_2021\Combined_Data\Extra\Statistics\Thijs

1.e

Pathway Soil Classification