READ ME File For 'Dataset for Monitoring macronutrient dynamics in soil and water with droplet microfluidic sensors' Dataset DOI: 10.5258/SOTON/D3782 ReadMe Author: James Lunn, University of Southampton ORCID ID 0000-0003-1961-3336 This dataset supports the thesis entitled Monitoring macronutrient dynamics in soil and water with droplet microfluidic sensors AWARDED BY: University of Southampton DATE OF AWARD: 2026 Thesis UUID: 536bccff-2858-4791-bada-04bc4b2345fa Date of data collection: Jan 2021 to Nov 2024 Information about geographic location of data collection: Laboratory Experiments (Development, Glucose Experiment, Phosphate Experiments, Bioreactor Experiment) - University of Southampton, Hampshire, UK Campus Deployment - University of Southampton, Hampshire, UK Writtle Forest Deployment - Writtle Forest, Essex, UK Organic Amendment Experiment - University of Reading, Berkshire, UK Licence: CC-BY 4.0 Related projects/Funders: Natural Environment Research Council - NE/T010584/1 --- DATA & FILE OVERVIEW --- This dataset contains: [File list (filenames, directory structure (for zipped files) and brief description of all data files)] Dataset for Thesis.xlsx - Dataset for each plotted figure in the thesis. This features one sheet for each graph of the thesis, providing the data used in each graph after processing with raw data if it is not provided in other files. Writtle Forest.zip - Collection of raw data files from the Writtle Forest experiment (3.3.4.2 in thesis). In this zip folder, the files are first divided by the time period they were collected in and then by analyser, with Sensor 1 being used for plots including Analyser A1 and A2 and Sensor 2 being used for Analyser B1 and B2. Sensor 3 (Also referred to as Sensor 5) collected data under a tree with Chronic Oak Decline, yet this was not featured in the thesis or in the related paper. Organic Amendment.zip - Collection of raw data files from the Organic Amendment experiment (6.1 in thesis). This folder contains a folder for each of the types of amendments - Control, Glycerol, and Straw - from which figures 52 to 54 of the thesis were derived.   High carbon organic amendment experiment collaboration UoR and UoS Nov 24.xlsx - Grab sample data (In units of mg/L N-NH3 or mg/L N-NO3). Performed and recorded by Ellie Barbrook (University of Reading, ORCID ID 0000-0002-0054-9146) Bioreactor.zip - Collection of raw data files for the bioreactor experiments (6.2 in thesis). This folder contains individual files for each experiment.   Figure 57 - 2024-08-21 CALIBRATION.csv - Raw calibration file from the multinutrient analyser for the Figure 57 - 2024-08-30 SYSTEM RUNNING 12 hour cycle 100 mL RAS.csv file.   Figure 57 - 2024-08-30 SYSTEM RUNNING 12 hour cycle 100 mL RAS.csv - The analyser data file for the Figure 57 experiment.   Figure 57 - Calibration TImings.xlsx - Timings for "Figure 57 - 2024-08-21 CALIBRATION.csv" in rows H-I. Other information regarding peristaltic pump flow rate in rows C-E yet this does not relate to the calibration.   Figure 58 - 2024-09-30 Calibration With Filter.csv - Filter run for figure 58, which was converted into concentration and the timing changed to line up with the non-filter run to make the graph.   Figure 58 - 2024-10-01 Calibration No FIlter.csv - Non-filter run for figure 58.   Figure 59 - 2024-10-10 System Running.csv - Analyser output for Figure 59     Figure 59 - 2024-10-11 WP Analysis - Folder containing the Well plate files for grab sample analysis for Ammonium, Nitrate and Nitrite.     Figure 59 - Analyser Calibration   2024-10-09 New Rgts Calibration.csv - Calibration file for "Figure 59 - 2024-10-10 System Running.csv".   2024-10-09 Calibration Timings.xlsx - Information about how "2024-10-09 New Rgts Calibration.csv" was run.     Figure 59 - WP Calibration - Folder containing the well plate files for the calibration of "2024-10-11 WP Analysis"   Figure 60 - Analyser Data - Folder containing the analyser data for both days of figure 60's plot.   06112024 analyzer data.xlsx - Day 1   07112024 analyzer data.xlsx - Day 2 Figure 31b - Glucose Test - Ultrafiltration.zip - Data for Figure 31b   2022-10-27 PD1 10 mM Control PD2 2p5 Glucose Calibration.csv - Calibration run of this experiment   2022-10-27 PD1 10 mM Control PD2 2p5 Glucose Test 27-10 to 31-10.csv - First 5 days of Ultrafiltration run   2022-10-27 PD1 10 mM Control PD2 2p5 Glucose Test 31-10 to 03-11.csv - Days 5 to 8 of Ultrafiltration run   2022-10-27 PD1 10 mM Control PD2 2p5 Glucose Test 03-11 to 04-11.csv - Days 8 to 9 of Ultrafiltration run Relationship between files, if important for context: "Dataset for Thesis.xlsx" includes processed data from each .zip file in the corresponding sheets for the figures. Additional related data collected that was not included in the current data package:   Raw data for the following figures was collected but not provided as I do not have access to the raw files (Collected by Dr. Bingyuan Lu):   Figure 24b   Figure 26   Figure 31 - Microdialysis   Figure 33d   Figure 35c and d (Writtle Forest)- Raw data from 14/09/2022 to 04/10/2022   Figure 36 (Writtle Forest)- Conversion of Grab Samples provided by Reading to what was included with the graph   Figure 40 (Phosphate)   Figure 41d (Phosphate) --- METHODOLOGICAL INFORMATION --- Description of methods used for collection/generation of data: Methods of experimental design and data collection for chapter 3 outlined in the paper "Droplet Microfluidic-Based In Situ Analyzer for Monitoring Free Nitrate in Soil" (Lu et al., 2024, https://doi.org/10.1021/acs.est.3c08207). Methods of experimental design and data collection for chapter 4 outlined in the paper "Highly sensitive absorbance measurement using droplet microfluidics integrated with an oil extraction and long pathlength detection flow cell" (Lu et al., 2024, https://doi.org/10.3389/fchem.2024.1394388). Primary method of data collection is through the collection of variables from various droplet microfluidic analysers developed in-house, outputting data in the form of intensity collected from photodiodes(TSL257, Texas Advanced Optoelectronic Solutions Inc.) inserted into a "flow cell", in which an LED using a specific wavelength of light shines light through a section of thin-walled PTFE tubing (UT7 tubing (0.7 mm ID)) through which the droplet microfluidic system flows, on the other side of this tubing is the photodiode. Light is absorbed from the colorimetric solution giving a difference in intensity, which is recorded. Temperature data is recorded from a thermistor built into the heater system. Soil moisture data is recorded and converted directly into a WHC% through the Arduino, as outlined in section 3.2.4 of the thesis. Grab sample absorbances for the Bioreactor experiment and the Glucose assay were collected using a FLUOstar Omega (BMG Labtech) plate reader. Measurements for Nitrate and Nitrite were taken at 540 nm using the Griess+ assay (http://dx.doi.org/10.1021/acs.est.9b01032), whilst Ammonium measurements were taken at 660 nm using a modified salicylate assay (Krom, M. D., Spectrophotometric determination of ammonia: a study of a modified Berthelot reaction using salicylate and dichloroisocyanurate. The Analyst 1980, 105 (1249), 305.). Methods for processing the data: Raw intensity data from the analysers was processed through use of a MATLAB script which detected oil and aqueous plateaus (determined by a threshold above and below the level of intensity for oil, which remains consistent if temperature is stable). This output would then produce a .csv file containing the average time of each aqueous droplet, the relative time to the start of the data file, and the averaged intensity for each aqueous droplet. This output file was then transferred to an OriginLab file associated to the experiment and analysed. The aqueous intensities could then be converted into an absorbance through the Beer-Lambert law (A = -log(I/I0)), with "A" being Absorbance, "I" being the intensity of the aqueous sample, and "I0" being the intensity of the 'blank', produced by flowing Milli-Q through the sample line of the analyser. Through creation of a calibration plot, in which several standard solutions of different concentrations are flowed through the sample line of the analyser and the absorbances from each of the standard solutions are plotted against their concentration, the absorbances can be converted into concentrations. Software- or Instrument-specific information needed to interpret the data, including software and hardware version numbers: MARS Data Analysis Software (Provided by BMG Labtech, Version 3.32) - Analysis of FLUOstar Omega data for grab sample analysis, including accounting for blank readings automatically. Standards and calibration information, if appropriate: Before each analyser run, a calibration curve was made through adding standard samples to the analyser to create a minimum of 60 droplets per sample (Half an hour of run time at 2 RPM) to allow for a plateau of intensity to form for each concentration. This plateau would then be averaged to give one value per concentration, with the standard deviation between these plateaued droplets forming the error. A blank was also run to allow for the conversion of these intensities into absorbances. Environmental/experimental conditions: For all studies performed in the laboratory, People involved with sample collection, processing, analysis and/or submission: University of Southampton: Prof. Xize Niu, Prof. Adrian Nightingale, Dr. Bingyuan Lu, Dr. Liam Carter, Jayshree Bunge, Ken Yeung, José Fabelo-Morales, Dr. Jelena Milinovic, Greg Slavik University of Reading: Prof. Liz Shaw, Dr. Selva Dhandapani, Ellie Barbrook --- DATA-SPECIFIC INFORMATION Dataset for Thesis.xlsx --- Figure 16b - Raw temperature data from heater unit in degrees C. Number of variables: 3 Number of cases/rows: 10002 Variable list, defining any abbreviations, units of measure, codes or symbols used: Time (Year-Month-Day Hours:Minutes:Seconds.Tenths of seconds), Time (seconds), Temperature (Celsius)   Missing data codes: NaN Date that the file was created: October, 2023 Figure 20: Raw data superimposed with processed droplet data for both microdialysis and ultrafiltration during a calibration early in the analyser's development. Number of variables: 11 Number of cases/rows: 209309 Variable list, defining any abbreviations, units of measure, codes or symbols used: Time, PD1 Intensity (Microdialysis), PD2 Intensity (Ultrafiltration), PD1 Droplet Number (DropletNumber), PD1 Absolute Time (Absolute_Time), PD1 Relative Time (Relative_Time), PD1 Droplet Intensity (Value), PD2 Droplet Number (DropletNumber), PD2 Absolute Time (Absolute_Time), PD1 Relative Time (Relative_Time), PD1 Droplet Intensity (Value   Missing data codes: NaT, NaN Date that the file was created: October, 2021 Figure 21 - Microdialysis probe placed in a standard solution and pumped via the 3 flow types (syringe pump, peristaltic pump (push) and peristaltic pump (push-pull) Absorbance data taken from UV-Vis spectrometer at 540 nm, 3 samples taken per flow rate and averaged, compared to that of the standard which the microdialysis probes were placed in. The pre-averaging data was not available for push-pull as I can not access this file. Number of variables: 4 Number of cases/rows: 16 Variable list, defining any abbreviations, units of measure, codes or symbols used: Flow Rate / µL/min, Relative Recovery / % (Syringe Pump), Relative Recovery / % (Push), Relative Recovery / % (Push/Pull) Missing data codes: NaN Date that the file was created: July, 2021 Figure 23 - Soil S5: Soil Moisture Calibration, Variables in order of Time and Soil1 were used in this calibration, with Soil 2 not being connected. 3 moisture readings were taken of each soil moisture level (40%, 50% ... 100%). These readings were then converted using equation 7 ( Moisture reading = (ADC(air) - ADC(soil)) / ADC(air) ) into normalised values, which were then plotted. The red value is a prediction of what the moisture content would be when the sensor was maxed out (approx. 120%) using the straight line response from the other values. Number of variables: 18 Number of cases/rows: 159989 Variable list, defining any abbreviations, units of measure, codes or symbols used: Time (Year-Month-Day Hours:Minutes:Seconds.Tenths of seconds), Bit Resolution (10-bit), ADC Average (how many unique digital values the ADC values can have, for PD Intensity and soil moisture), PD1 Intensity (ADC Value received from the photodiode in flow cell 1), PD2 Intensity, PD3 Intensity, PD4 Intensity, Brightness1 (Brightness set for the LED in flow cell 1), Brightness2, Brightness3 , Brightness4, Soil 1 (ADC value from soil moisture probe 1), Soil 2, Heater 1 (Heater temperature from Heater 1 in degrees C), Heater 2, Motor 1 Speed (Revolutions per minute of motor 1), Motor 2 Speed, Motor 3 Speed. Date that the file was created: November, 2022 Figure 24 - Soil S1: Calibration of the sensor at low concentrations (0.5 to 2 mM). The 16-bit droplet intensities were averaged and converted into absorbances using the beer-lambert law (A=-log(I/I0)), which were then plotted against concentration to create the calibration curve. I am unable to find the raw data for this, yet I have found processed data that was used for the creation of the soil paper figure. Figure 25: Solution calibrations for microdialysis and ultrafiltration. Variables in order of PD1 is microdialysis, PD2 is ultrafiltration. This raw data was then passed through a matlab script to find aqueous droplet intensity and then the average from each concentration was plotted. The processed data is in columns G-K, with H and I giving MD and UF Intensity, which is then converted into absorbance for J and K, which was then plotted. Number of variables: 5 Number of cases/rows: 209308 Variable list, defining any abbreviations, units of measure, codes or symbols used: Time, PD1 Intensity, PD2 Intensity, Heater 1 / C, Heater 2 / C. Missing data codes: NaN Date that the file was created: October, 2021 Figure 26: MD probe connected to the analyser was placed in some "blanked then doped" soil (soil that was rinsed out wih water to remove nitrate present, then dried, then doped to a specific water content using a standard solution) then run intermittently, varying the stop times between runs. From this, raw data was then processed by manually picking out the intensities from the microdialysis peaks from each interval, which were then converted to absorbances, then plotted with time. I did not have access to the raw data files as these were made and processed by Dr. Bingyuan Lu, who shared the final analysed file with me only showing the intensities of the peaks and their respective absorbances. Figure 27: Calibrations run from each different WHC% level (60, 70, 84, 100 %WHC). 27a and 27b include calibrations that were taken with little stopping time between each sample taken; after 30 minutes in each sample the motors were stopped whilst the probes were placed into a different sample. Columns A to R show the processed data used in each graph. Columns S to AS have the raw data for graphs 27a and 27b, with intensities vs time with heater data (40 C throughout each experiment). Columns AU to BG show the raw intensity data for figure 27c, which show the intensity recorded for a 12 hour gap and then a 6 hour gap afterwards. Number of variables: 5 Number of cases/rows: 209308 Variable list, defining any abbreviations, units of measure, codes or symbols used: Time, PD1 Intensity, PD2 Intensity, Heater 1 / C, Heater 2 / C. Missing data codes: NaN Date that the file was created: December, 2021 to January, 2022 Figure 29: Continuous monitoring of soil through wetting, showing how each method responds to a change in moisture. A microdialysis probe and an ultrafiltration probe were inserted into sieved dry soil, which was gently packed down. The analyser connected to the probes was then turned on. After a short period, water was then added to wet the soil up to 100% WHC. Columns A+B for microdialysis, C+D for ultrafiltration, F to J for the raw data from the sensor. Number of variables: 5 Number of cases/rows: 1046812 Variable list, defining any abbreviations, units of measure, codes or symbols used: Time, PD1 Intensity, PD2 Intensity, Heater 1 / C, Heater 2 / C. Missing data codes: NaN Date that the file was created: October, 2022 Figure 30: Continuous monitoring of doped soil with glucose added, columns A+B for glucose added and columns C+D for control. Columns F to I for the processed droplet data, giving the intensity of aqueous droplets. Number of variables: 4 Number of cases/rows: 43321 Variable list, defining any abbreviations, units of measure, codes or symbols used: Time, PD1 Intensity, Time, PD2 Intensity. Missing data codes: NaN Date that the file was created: March, 2022 Figure 31 - Soil 4: a) Change of dialysate nitrate (%) versus experimental day using microdialysis using different doped glucose concentrations. b) Change of pore water nitrate (%) versus experimental day using ultrafiltration using different doped glucose concentrations. Lower nitrate levels were used for the ultrafiltration study as the ultrafiltration probes collected 100% of the nitrate concentration in pore water as opposed to approx.. 30% collected through using microdialysis. Raw data for the glucose assay is available in the zip folder "Figure 31b - Glucose Test - Ultrafiltration.zip" in the column order (Year-Month-Day Hours:Minutes:Seconds.Tenths of seconds), Bit Resolution (10-bit), ADC Average (how many unique digital values the ADC values can have, for PD Intensity and soil moisture), PD1 Intensity (ADC Value received from the photodiode in flow cell 1), PD2 Intensity, PD3 Intensity, PD4 Intensity, Brightness1 (Brightness set for the LED in flow cell 1), Brightness2, Brightness3 , Brightness4, Soil 1 (ADC value from soil moisture probe 1), Soil 2, Heater 1 (Heater temperature from Heater 1 in degrees C), Heater 2, Motor 1 Speed (Revolutions per minute of motor 1), Motor 2 Speed, Motor 3 Speed. ; raw data for the microdialysis assay is unavailable as this was recorded by Dr. Bingyuan Lu and I have access only to the processed data for this. Figure 32 - Plate reader analysis of the glucose assay. Columns B+C for Glucose, D+E for Control. Columns H, I , L and M are the raw absorbances from the plate reader, which were averaged into columns J and N with standard deviations in K and O. The calibration from the plate reader is in rows P to R, with the samples being diluted by a factor of 50 to get them into the range for the Griess+ assay (30 mM expected peak, Griess assay starts to plateau after 800 micromolar) Figure 33d - Soil 5: Campus deployment of soil analyser; Pore water nitrate for ultrafiltration, Dialysate nitrate for microdialysis. I do not have access to the raw data as this was processed by Dr. Bingyuan Lu and only the processed data was shared with me. Number of variables: 3 Number of cases/rows: 49 Variable list, defining any abbreviations, units of measure, codes or symbols used: Days, Measured Nitrate / mM (Pore water nitrate), Measured Nitrate / mM (Dialysate nitrate) Missing data codes: NaN Date that the file was created: September, 2022 Figure 35c and d - Soil 6: 2 month deployment of the sensors at Writtle, using two microdialysis analysers (A and B) underneath different oak trees. A1 and B1 were installed using the rectangular groove method, A2 and B2 were installed using the needle method, which had worse soil contact. Columns A to L in the Dataset show recovered nitrate (mM) before conversion into micrograms of NO3 per gram of dried soil. The raw data files recorded every 6 hours of the deployment for the analysers are in the Grouped Writtle Visits folder, separated first by the time period they were collected in (Each folder represents a maintenance event, on which data was retrieved from the SD card of each device) then these were separated by which sensor they were collected by. Sensor 1 represented Analyser A1 and A2, with PD1 being A1 and PD2 being A2, Sensor 2 represented Analyser B1 and B2, with PD1 being B1 and PD2 being B2, while Sensor 3 was not represented in graphs. This analyser was placed next to a tree with Chronic Oak Decline that we had tried to use homemade microdialysis probes which did not give consistent enough data to be published alongside the others. Each file has a timestamp of when the file was created, with earlier files named via DD-MM-hh-mm convention e.g. 19-Oct-07-58; this was later changed in Jan 2023 to give a more typical file format of YYYY-MM-DD hh-mm to allow for easier processing, as the files would be in ascending order with time if listen alphabetically, making it easier for MATLAB to interpret. To process these raw files, these were displayed on MATLAB and the peak water droplet value and a corresponding oil value from each raw data file was recorded manually and put into a .csv file. This aqueous intensity was then converted into a concentration (millimolar) using a calibration recorded during each maintenance period. Figure 35 uses the first two months of data, from 14/09/2022 to 14/11/2022. Raw data from 14/09/2022 to 04/10/2022 is missing as this was handled by Dr. Bingyuan Lu. Rainfall data was collected from a nearby data center (Chelmsford Weather Station). Number of variables: 10 Number of cases/rows: 640 Variable list, defining any abbreviations, units of measure, codes or symbols used: Experimental Time / Days, Rainfall / mm, Soil Moisture / %WHC, Recovered nitrate / mM (Analyser A1), Recovered nitrate / mM (Analyser A2), Experimental Time / Days, Rainfall / mm, Soil Moisture / %WHC, Recovered nitrate / mM (Analyser B1), Recovered nitrate / mM (Analyser B2) Missing data codes: NaN Date that the file was created: September, 2022 to October, 2023 Figure 36 - The concentrations from Figure 35 were converted into micrograms of nitrate per gram of dried soil through dividing the concentration (in micromolar) by the relative recovery of the probe at the moisture content of the soil, then multiplying this by the molecular weight of Nitrate (62.0049 g/mol) and by the moisture content (%WHC), which was recorded from the moisture probe, with the moisture value from each pickup being averaged from the last 1000 data points. Grab sample data provided by University of Reading; whilst I have access to the grab samples in mg/L I do not have access to the conversion of this data to the grab samples in the processed graph as this was done by Dr. Bingyuan Lu. Number of variables: 10 Number of cases/rows: 641 Variable list, defining any abbreviations, units of measure, codes or symbols used: Day, Absolute soil nitrate (µg NO3- per g of dried soil, Location 1), Absolute soil nitrate (µg NO3- per g of dried soil, Location 2), S1 WHC (%), S2 WHC (%), Day, Grab Sample Concentration (µg NO3- per g of dried soil, Location 1), Error, Grab Sample Concentration (µg NO3- per g of dried soil, Location 2), Error Missing data codes: NaN Date that the file was created: September, 2022 to October, 2023 Figure 37 - Two graphs of the year-long deployment. For a, columns B to D for analyser 1 data, E for rainfall, F and G for grab sample data. For b, columns I to K for analyser 2 data, L for rainfall, M and N for grab sample data. Day 0 is 14/09/2022 for both analysers. Number of variables: 12 Number of cases/rows: 838 Variable list, defining any abbreviations, units of measure, codes or symbols used: Day,S1 Soil nitrate (µg NO3- per g of dried soil, Location 1), S1 WHC (%, Location 1), Rainfall, Grab Sample Day, Grab Sample S1 (µg NO3- per g of dried soil, Location 1). Day,S2 Soil nitrate (µg NO3- per g of dried soil, Location 2), S2 WHC (%, Location 2), Rainfall, Grab Sample Day, Grab Sample S2 (µg NO3- per g of dried soil, Location 2) Missing data codes: NaN Date that the file was created: September, 2022 to October, 2023 Figure 40 - Phosphate 2: a) Effective pathlength versus theoretical pathlength for the flow cells. Columns B and C. b) Taylor dispersion measurement for the different flow cells. Columns E to J for the long pathlength flow cells (5 to 20 mm), columns K and L for UT7 flow cell. This data was collected by Dr. Bingyuan Lu and I do not have access to the raw data. Figure 41a, b, c - Phosphate 3: a) Shows the raw data with the determined droplet values superimposed, whilst b) shows a zoomed in section. The raw data for a) and b) is present in columns B and D (PD2). The processed aqueous droplet value information is present in columns F to H, with the raw droplet data being divided by 4 (going from 12-bit to 10-bit), hence why the oil droplet level is at around 3600 in the raw data. c) shows the filtering of data, in which the data is averaged every 5 points. The raw data for this averaging is present in columns J and K. Number of variables: 3 Number of cases/rows: 89158 Variable list, defining any abbreviations, units of measure, codes or symbols used: Time, PD1 Intensity, PD2 Intensity Date that the file was created: November, 2023 Figure 41d - Phosphate 3: Calibration graphs for each of the flow cells from 2.5 to 10 micromolar, compared to a UV spectrometer. This data was recorded by Dr. Bingyuan Lu and I do not have access to the raw data. Figure 42b - Phosphate 4: River sample analysis using the flow cell, compared to UT7 flow cell and a UV spectrometer. Columns B and C for the 20 mm flow cell, D and E for the UT7 flow cell, and F and G for the UV spectrometer. I to Q show the raw data, in which a blank and 5 mM standard was run before the river samples, which were run in triplicate and in order from 1 to 7. Number of variables: 9 Number of cases/rows: 437449 Variable list, defining any abbreviations, units of measure, codes or symbols used: Time, PD1 Intensity, PD2 Intensity, Soil Moisture Reading, Soil Moisture %, Heater 1 (Celsius), Heater 2 (Celsius), Box Temperature (Celsius), Box Humidity (%). Soil Moisture Reading, Soil Moisture (%), Heater 1 and Heater 2 were not used in this experiment. Date that the file was created: November, 2023 Figure 47 - Interference study for the 3 flow cells, showing the lack of change of the intensity to changing the brightness of other LEDs of the other flow cells. Columns A and B for PD1, C and D for PD2, E and F for PD3. Number of variables: 6 Number of cases/rows: 3745 Variable list, defining any abbreviations, units of measure, codes or symbols used: Time, PD1 Intensity, Time, PD2 Intensity, Time, PD3 Intensity. Date that the file was created: January, 2024 Figure 48 - Heater response from activation of the unit, showing very little change to the heating of the system with the presence of the aluminium rod. Columns C and D used for the graph. Number of variables: 6 Number of cases/rows: 12001 Variable list, defining any abbreviations, units of measure, codes or symbols used: Recorded at 20 Hz (Row Number), Time / s, Time / minutes, Heater Temperature / Degrees C Date that the file was created: July, 2024 Figure 49 - Comparison of Ammonium Calibration between the droplet analyser and a well-plate reader for the organic amendment assay. Raw Data is from the droplet analyser. Number of variables: 4 Number of cases/rows: 542 Variable list, defining any abbreviations, units of measure, codes or symbols used: Droplet Number, Time, Relative Time, Intensity Missing data codes: NaT, NaN Date that the file was created: February, 2024 Figure 50: Calibration curves for the different nutrients, with columns A-C used for NH4, D-F for NO2 (PD2), G-I for NO2 (PD3), and J-L for NO3. NO2 is calibrated in both PD2 and PD3 to allow for the discernment of NO3 in PD3, as both NO2 and NO3 produce colour in the PD3 line due to the presence of vanadium (III) chloride, which reduces NO3 to NO2. The calibration took two sessions to complete due to the slow flow of the analyser (2 RPM). The droplet data for both days' calibration is in columns A to AD. Columns AF to AV contain averages to the intensity data for each concentration, which is then converted to absorbance to build up calibration curves. Columns AX to BA give information on the blanks for each photodiode line, as well as the gradients and intercepts for each calibration trendline. Columns BC to BE show the order of operations of each calibration, identifying the order and timings of each sample. - first day calibration PD1_Output Number of variables: 5 Number of cases/rows: 642 Variable list, defining any abbreviations, units of measure, codes or symbols used: Droplet Number, Time, Relative Time, Intensity, Absorbance Missing data codes: NaT, NaN Date that the file was created: October, 2024 - first day calibration PD2_Output Number of variables: 5 Number of cases/rows: 666 Variable list, defining any abbreviations, units of measure, codes or symbols used: Droplet Number, Time, Relative Time, Intensity, Absorbance Missing data codes: NaT, NaN Date that the file was created: October, 2024 - first day calibration PD3_Output Number of variables: 5 Number of cases/rows: 1206 Variable list, defining any abbreviations, units of measure, codes or symbols used: Droplet Number, Time, Relative Time, Intensity, Absorbance Missing data codes: NaT, NaN Date that the file was created: October, 2024 - second day calibration PD1_Output Number of variables: 5 Number of cases/rows: 553 Variable list, defining any abbreviations, units of measure, codes or symbols used: Droplet Number, Time, Relative Time, Intensity, Absorbance Missing data codes: NaT, NaN Date that the file was created: November, 2024 - second day calibration PD2_Output Number of variables: 5 Number of cases/rows: 594 Variable list, defining any abbreviations, units of measure, codes or symbols used: Droplet Number, Time, Relative Time, Intensity, Absorbance Missing data codes: NaT, NaN Date that the file was created: November, 2024   - second day calibration PD3_Output Number of variables: 5 Number of cases/rows: 1079 Variable list, defining any abbreviations, units of measure, codes or symbols used: Droplet Number, Time, Relative Time, Intensity, Absorbance Missing data codes: NaT, NaN Date that the file was created: November, 2024 Figures 52 - 54: All 3 figures use the data shown here, with 54 being an overview of the entire period. Column A gives the time for all analyser data. Columns B-H show the analyser data in mg/L N and columns I-N show the analyser data in µg N /g dried soil. Grab sample data in mg/L N is shown in columns P-V and is shown in µg N /g dried soil in columns X-AD. The "Organic Amendment.zip" file contains the raw data from each of the three analysers, with PD1 and PD2 being replicates. - Analyser Data Number of variables: 13 Number of cases/rows: 110 Variable list, defining any abbreviations, units of measure, codes or symbols used: Date, Concentration (mg/L N, NH4-N Control), Concentration (mg/L N, NO3-N Control) Concentration, (mg/L N, NH4-N Glycerol), Concentration (mg/L N, NO3-N Glycerol), Concentration (mg/L N, NH4-N Straw), Concentration (mg/L N, NO3-N Straw), Concentration (µg N /g dried soil, NH4-N Control), Concentration (µg N /g dried soil, NO3-N Control), Concentration, (µg N /g dried soil, NH4-N Glycerol), Concentration (µg N /g dried soil, NO3-N Glycerol), Concentration (µg N /g dried soil, NH4-N Straw), Concentration (µg N /g dried soil, NO3-N Straw) Missing data codes: NaT, NaN Date that the file was created: February, 2024 to August, 2024 - Grab Sample Data Number of variables: 14 Number of cases/rows: 14 Variable list, defining any abbreviations, units of measure, codes or symbols used: Date, Concentration (mg/L N, NH4-N Straw), Concentration (mg/L N, NO3-N Straw), Concentration, (mg/L N, NH4-N Glycerol), Concentration (mg/L N, NO3-N Glycerol), Concentration (mg/L N, NH4-N Control), Concentration (mg/L N, NO3-N Control), Date, Concentration (µg N /g dried soil, NH4-N Straw), Concentration (µg N /g dried soil, NO3-N Straw), Concentration, (µg N /g dried soil, NH4-N Glycerol), Concentration (µg N /g dried soil, NO3-N Glycerol), Concentration (µg N /g dried soil, NH4-N Control), Concentration (µg N /g dried soil, NO3-N Control) Date that the file was created: February, 2024 to August, 2024 Figure 57: Droplet data for the first bioreactor run. Columns A and B for Ammonium, C+D for Nitrate+Nitrite, E+F for Nitrite. Grab sample times are in column G. Grab sample Ammonium is in columns H+I, Grab sample Nitrate+Nitrite in columns J+K, Grab sample nitrite in columns L+M. Raw data for the grab sample analysis is in columns O to S, with nitrate being calculated from columns R and S. The raw data for the bioreactor run is in the file "Figure 57 - 2024-08-30 SYSTEM RUNNING 12 hour cycle 100 mL RAS.csv" in the "Bioreactor.zip" folder. The order of variables in this file is "Time", "PD1 Intensity", PD2 Intensity", "PD3 Intensity", "Soil Moisture - Raw", "Soil Moisture - %", "Heater 1 Temperature", "Heater 2 Temperature", "Box Temperature", "Box Humidity". The soil moisture probe and heater 2 were not connected for the bioreactor experiments. This raw data was then converted by calculating droplet intensity as before, alongside a calibration beforehand this was then converted into the concentrations shown in columns A-F of the dataset sheet. - Droplet Data Number of variables: 6 Number of cases/rows: 17084 Variable list, defining any abbreviations, units of measure, codes or symbols used: Time, Concentration (mM, Ammonium), Time, Concentration (mM, Nitrate + Nitrite), Time, Concentration (mM, Nitrite) Missing data codes: NaT, NaN Date that the file was created: August, 2024 to September, 2024 Grab Sample Data - Columns O to T Number of variables: 6 Number of cases/rows: 50 Variable list, defining any abbreviations, units of measure, codes or symbols used: Identity, Time of Sampling, Absorbance - Ammonium, Absorbance - Nitrate+Nitrite, Absorbance - Nitrite, Absorbance - Nitrate Missing data codes: NaT, NaN Date that the file was created: September, 2024 Figure 58: Comparison in sample changeover time with and without the resin filter. Raw time for the no filter run in column A. This is converted into minutes for column B, with the absorbance for no filter in column C. Column D shows the time in seconds for the filter (negative as it lines up with a section of no-filter) and this is converted to minutes in column E with a second correction factor to the part shown in the graph. Absorbance for the filter is in column F. Columns B+C and E+F are used for the no filter and filter plots respectively. Raw intensity data for the filter and no filter runs are in "Figure 58 - 2024-09-30 Calibration With Filter.csv" and "Figure 58 - 2024-10-01 Calibration No FIlter.csv" in the "Bioreactor.zip" folder. The order of variables in these files are "Time", "PD1 Intensity", PD2 Intensity", "PD3 Intensity", "Soil Moisture - Raw", "Soil Moisture - %", "Heater 1 Temperature", "Heater 2 Temperature", "Box Temperature", "Box Humidity". Timings for each sample are in columns H to M of the "Figure 58" dataset sheet. Number of variables: 6 Number of cases/rows: 1452 Variable list, defining any abbreviations, units of measure, codes or symbols used: Time (Seconds, No Filter), Time for Graph (Minutes, Filter), Absorbance (Filter), Time (Seconds, Filter), Time, (Minutes, Filter), Absorbance (Filter) Missing data codes: #VALUE! Date that the file was created: October, 2024 Figure 59: Bioreactor run after implementing the resin filter. Columns A-G for the analyser data, H-L for the grab sample data. Raw data for the experiment is in the file "Figure 59 - 2024-10-10 System Running.csv" in the "Bioreactor.zip" folder. The order of variables in these files are "Time", "PD1 Intensity", PD2 Intensity", "PD3 Intensity", "Soil Moisture - Raw", "Soil Moisture - %", "Heater 1 Temperature", "Heater 2 Temperature", "Box Temperature", "Box Humidity". Calibration and timings for the analyser are in the "Figure 59 - Analyser Calibration" folder in "Bioreactor.zip". Well plate calibration and analysis are in the "Figure 59 - WP Calibration" and "Figure 59 - 2024-10-11 WP Analysis" folders respectively in "Bioreactor.zip". Number of variables: 6 Number of cases/rows: 50 Variable list, defining any abbreviations, units of measure, codes or symbols used: Experiment Time, Concentration (mM, Ammonium), Experiment Time, Concentration, (mM, Nitrate), Experiment Time, Concentration (mM, Nitrite), Concentration (mM, Total), Grab Sample Time, Concentration (mM, Ammonium Grab), Concentration (mM, Nitrate Grab), Concentration (mM, Nitrite Grab), Concentration (mM, Total Grab) Missing data codes: NaT, NaN Date that the file was created: October, 2024 Figure 60: Two day feedstock only run. Bar chart showing ammonium, nitrite and nitrate and their respective concentrations over the two days. Columns B+C for Day 1, D+E for Day 2. Calibration for this file is in the sheet "Figure 50" Number of variables: 5 Number of cases/rows: 6 Variable list, defining any abbreviations, units of measure, codes or symbols used: Chemical, Concentration (mM, Day 1), Error, Concentration (mM, Day 2), Error Missing data codes: NaT, NaN Date that the file was created: October, 2024 --- DATA-SPECIFIC INFORMATION - Writtle Forest.zip --- Number of variables: 9 Number of cases/rows: 36001 Variable list, defining any abbreviations, units of measure, codes or symbols used: Time (DD/MM/YYYY HH:mm:ss.0), PD1 Intensity(PD1AVG), PD2 Intensity(PD2AVG), Raw Moisture Sensor reading (MoistureSensor), Processed moisture sensor reading (%, MoistureSensor(%)), Heater 1 Temperature (Degrees C, Heater1Temp), Heater 2 Temperature (Degrees C, Heater2Temp), Box Temperature (Degrees C, Box_Temp(C)), Box Humidity (%, Box_Hum(%))   Missing data codes: NaN Date that the file was created: November, 2022 to October, 2023 --- DATA-SPECIFIC INFORMATION Figure 31b - Glucose Test - Ultrafiltration.zip - Data for Figure 31b --- Number of variables: 18 Number of cases/rows: Variable list, defining any abbreviations, units of measure, codes or symbols used: Time (Year-Month-Day Hours:Minutes:Seconds.Tenths of seconds), Bit Resolution (10-bit), ADC Average (how many unique digital values the ADC values can have, for PD Intensity and soil moisture), PD1 Intensity (ADC Value received from the photodiode in flow cell 1), PD2 Intensity, PD3 Intensity, PD4 Intensity, Brightness1 (Brightness set for the LED in flow cell 1), Brightness2, Brightness3, Brightness4, Soil 1 (ADC value from soil moisture probe 1), Soil 2, Heater 1 (Heater temperature from Heater 1 in degrees C), Heater 2, Motor 1 Speed (Revolutions per minute of motor 1), Motor 2 Speed, Motor 3 Speed.   Missing data codes: NaN Date that the file was created: October, 2022 --- DATA-SPECIFIC INFORMATION Organic Amendment.zip --- Number of variables: 9 Number of cases/rows: 36001 Variable list, defining any abbreviations, units of measure, codes or symbols used: Time (DD/MM/YYYY HH:mm:ss.0), PD1 Intensity(PD1AVG), PD2 Intensity(PD2AVG), Raw Moisture Sensor reading (MoistureSensor), Processed moisture sensor reading (%, MoistureSensor(%)), Heater 1 Temperature (Degrees C, Heater1Temp), Heater 2 Temperature (Degrees C, Heater2Temp), Box Temperature (Degrees C, Box_Temp(C)), Box Humidity (%, Box_Hum(%))   Missing data codes: NaN Date that the file was created: February, 2024 - June, 2024 --- DATA-SPECIFIC INFORMATION Bioreactor.zip --- Number of variables: 10 Variable list, defining any abbreviations, units of measure, codes or symbols used: Time (DD/MM/YYYY HH:mm:ss.0), PD1 Intensity(PD1AVG), PD2 Intensity(PD2AVG), PD3 Intensity(PD3AVG) Raw Moisture Sensor reading (MoistureSensor), Processed moisture sensor reading (%, MoistureSensor(%)), Heater 1 Temperature (Degrees C, Heater1Temp), Heater 2 Temperature (Degrees C, Heater2Temp), Box Temperature (Degrees C, Box_Temp(C)), Box Humidity (%, Box_Hum(%)) - Heater 2 and Moisture sensor were not used in these experiments.   Missing data codes: NaN Date that the file was created: August, 2024 - November, 2024