EART60061 Cheatsheet · Measuring and Predicting 1

The University of Manchester · open-book exam sheet for the final exam. Online open-book test (2 × 1.5 hours) after teaching week 12, Semester 1 2026/27 (check the date in your exam timetable). Full EART60061 study guide →

A4 landscape · two sides · six columns · print at 100%EART60061 · Manchester
EART60061
Measuring and Predicting 1
The University of Manchester · Postgraduate · Semester 1 2026/27
EART60061 cheatsheet QR code – open the full EART60061 study guide
Full guide + ask Sia
about EART60061
SIDE 1/2 METHODS · RULES · DECISIONSOPEN-BOOK EXAM SHEETOnline test · 100% · v2 · 2026-10-19

0 · Exam Mapcourse + Canvas

ONLINE TEST 100%

“The test will be open book and will draw on course materials and the exercises/tasks you are asked to complete throughout the unit.” (EART60061 unit page, Canvas, 2026/27 1st semester)

Format in the course data: multiple-choice + short-answer, covering every intended learning outcome, in two 1.5-hour sessions. The closest preparation is to rerun the exercises, change a parameter and explain the output.

If asked to…First move
quality of valuess → SE → t·SE
combine valueschoose error rule
test a liner² → t → p
predict processstate model + check
step a modelold values only
Exam move Name the method, substitute with units, carry the calculation to a decision, then state the assumption or check.

1 · Research ArgumentCh 1

generalintroAIMspecificintro3–6objectivesbroad → narrow → difficulty → tasksmethod organised by objectives
The written argument: the introduction earns the aim; difficulties earn ordered objectives.

quality testResearch = importance × novelty; rigour decides whether it counts.

Each objective must be achievable, recognisable and visibly linked to the aim. Order closed → open. Judge every section for coherence, completeness and concision. A Gantt chart should show parallel work and leave time for review.

PartMarker-facing test
General introductionbroad-to-narrow case ending at the requirement
Aimsingle statement, close to title
Specific introductionwhy the aim is difficult; terms defined
Objectives3–6 recognisable tasks, closed-to-open
Methodorganised by objectives; replication and error explicit
Planparallel activities, feasible timing, review allowed

Scope honestly: an MSc project need not have global importance or novelty; state whether each is local, national or wider. Rigour appears in suitable methods, replication and quantified uncertainty.

2 · Spread: Choose the StatisticCh 2

MeasureUse / warning
Rangequick; only min/max, hostage to one extreme
Mean absolute deviationsign-free; no distribution assumption
Sample suses every value; links spread to probability
Population σcomplete population only; divisor n

three spread formulasx̄ = Σxᵢ/n
MAD = Σ|xᵢ−x̄|/n
s = √[Σ(xᵢ−x̄)²/(n−1)]

Why n−1? The mean was estimated from the same sample, costing one degree of freedom. Replicate field/lab data: STDEV or STDEV.S; STDEVP divides by n. Signed deviations sum to zero; squaring removes sign and weights extremes.

3 · Normal DistributionCh 3

−1.96−1μ+1+1.9668.3%95% total
μ locates the peak; s sets width. 68.3% lies within ±1s and 95% within ±1.96s.

standardised = (x−x̄)/s · 95% range = x̄ ± 1.96s
within ±d: 1.00 → 68.3%; 1.96 → 95%; 2.58 → 99%

For a threshold, sketch first: one tail or both? Tail estimates assume random variation and that sample x̄,s represent the population. A home-built area table needs frequency × class width; coarse classes create resolution error.

dCentral shareBoth tails
1.0068.3%31.7%
1.9695.0%5.0%
2.5899.0%1.0%

Spreadsheet mechanics that get tested

  1. Set calculation to automatic before trusting changing outputs.
  2. For a frequency distribution, select all output cells, enter the data range and class limits, then use the array entry method shown in the exercise.
  3. Fix the mean cell with an absolute reference before copying deviation formulas.

4 · Confidence LimitsCh 4

1.96·SD1.96·SEt·SEuse for meansmall n widens t
Two corrections oppose: SE narrows from single readings to a mean; small-sample t widens.

four-line methodSE = s/√n
df = n−1
95% CL = t₀.₀₅,df × SE
report x̄ ± CL, units; interval [x̄−CL, x̄+CL]

Excel: TINV(0.05,n−1). Examples from the guide: df=2 → t=4.303; df=9 → 2.262; t approaches 1.96 only as n becomes very large. For two means, non-overlap supports a difference at 95%; overlap means do not claim a difference.

Error bar belongs toQuantity
single readingss or d·s
mean of nSE=s/√n
small-n 95% meant·SE

Quote mean and limit to the same decimal places. Figure legend must state 95% confidence. In sampling design, control tide, location, species/size and handling; replication measures variation that remains.

4b · Decision TrapCh 4 FAQ

SD describes measurements; SE describes their mean. Using 1.96·SD for a mean is too wide; using 1.96·SE with three replicates is too narrow because t=4.303.

Exam line “At 95% confidence the intervals [do/do not] overlap; therefore a difference [cannot/can] be claimed.”
ndf95% tUse
324.303triplicate mean
433.182four replicates
652.571six replicates
1092.262ten replicates
∞∞1.960normal limit

Biomonitoring design → defensible bars

Seaweed integrates exposure and concentrates metals, avoiding seawater's tide/current variation and salt matrix effects. Hold constant tide stage, sampling position, species and specimen size; replicate within each site. If a substitute species is unavoidable, sample both at a shared site and disclose the correction-factor ratio.

5 · Error PropagationCh 5

width ×depth× velocity= Q× conc.= flux× 86,400per dayexact multiplier keeps fractional error
Carry units first, then fractional errors through every measured multiplication.

select by operationsum/difference: E = √(Eₐ²+Eᵦ²+…)
product/quotient: E_Q/Q = √[(Eₐ/A)²+(Eᵦ/B)²+…]
power xⁿ: E/xⁿ = |n|·Eₓ/x

Assumptions: errors are random and uncorrelated. Direct addition overstates error at the stated confidence. An exact multiplier has zero error: it scales absolute error, not fractional error. The largest fractional input error usually dominates—identify it to say which measurement should improve.

Unit chainResult
m³/s × mg/Lg/s (1000 L/m³; 1000 mg/g)
g/s × 86,400 s/dayg/day
g/day ÷ 1000kg/day

5b · Flux & SamplingCh 5 case

springneaphours through tidal cycle
Flux can reverse with the tide; sample complete spring and neap cycles, not isolated concentrations.

Concentration is mass/volume; flux is mass/time and predicts transport. Estuary flux also varies through dilution, pH, adsorption, flocculation, settling, storms and stored metal. A negative tidal-cycle flux can mean net landward movement, not an arithmetic error. Compare a consent with the whole uncertainty interval.

ProcessDesign response
Dilutionmeasure discharge
pH / adsorptionanalyse dissolved + particulate
Settling / storageallow for later resuspension
Tidecomplete spring + neap cycles
Stormacknowledge rare unsampled export

Consent: report the result as a confidence interval; if it crosses the limit, the centre alone cannot settle compliance. Negative net flux during one spring tide can indicate temporary landward storage, not the long-run state.

Sense check Exact seconds/day leave fractional error unchanged.

6 · Modelling ProcessCh 6

real worldconceptualmathematicaloutputabstractformulatesolvetest
Numerical solution → verify; tune plausible parameters → calibrate; compare with observations → validate.
StageQuestion
Abstraction / assumptionWhat is omitted or simplified?
FormulationHow is the concept written mathematically?
SolutionExact analytical or stepped numerical?
VerificationDoes smaller Δt/Δx materially change it?
CalibrationWhich plausible parameter values fit?
ValidationDoes output match independent reality?

Verification ≠ validation. Verification checks numerical approximation; validation tests assumptions against observations. Analytical solutions need no numerical verification. Add complexity only when failed validation identifies a missing process.

7 · Regression & TransformationsCh 7

C=aQᵇlinear reaches 0 ✕CQ
Dilution is a power law: a straight line falsely reaches zero concentration.

significance from r²r = √r² · df = n−2
t = r√(n−2)/√(1−r²)
p = TDIST(t, n−2, 2); p<0.05 → significant at 95%

Report r² with p. Fit strength and significance answer different questions. Power law y=axᵇ → double-log straight line; exponential → semi-log. Splitting rising/falling hydrograph limbs captures hysteresis but reduces n, so recompute p.

ResultRead it as
large r², p≥.05fit without enough evidence
small r², p<.05real but weak relation

8 · Growth ModelsCh 8

Kexponentiallogistictime
Logistic growth follows exponential early, then the moderating term forces N toward K.

exponentialdN/dt = rN · Nₜ = N₀eʳᵗ · lnNₜ = lnN₀ + rt

logisticdN/dt = rN(K−N)/K
Nₜ = KN₀ / [N₀+(K−N₀)e⁻ʳᵗ]

Read N₀ and plateau K from data. While N≪K, r=(lnN₂−lnN₁)/(t₂−t₁). Natural logs match e; multiply a base-10 gradient by 2.303. If calibration fails, revisit assumptions.

9 · CompartmentsCh 9

X₁ litterX₂ soilF₀₁F₀₂F₁₀F₂₀F₁₂=K₁₂X₁
Donor-controlled flux: Fᵢⱼ=KᵢⱼXᵢ. Write each store as inputs minus outputs.

X₁′ = X₁ + Δt[F₀₁−(K₁₀+K₁₂)X₁]
X₂′ = X₂ + Δt[F₀₂+K₁₂X₁−K₂₀X₂]

Use old X₁ and old X₂ on both right-hand sides. Equilibrium: set changes to zero; X₁*=F₀₁/(K₁₀+K₁₂), X₂*=(F₀₂+K₁₂X₁*)/K₂₀. Residence time = 1/ΣKout. Check K·Δt<1; halve Δt and compare.

Output clueAction
logistic still missesrevisit model structure
store below X* and risingsystem is a sink
ΣKout·Δt>1refine the step
Δt/2 changes resultnot verified

10 · Fick Difference StepCh 10

Fᵢ→ᵢ₊₁ = Kd(Cᵢ−Cᵢ₊₁)
Cᵢ(t+1)=Cᵢ(t)+Kd[Cᵢ₋₁−2Cᵢ+Cᵢ₊₁]

Check total mass each step. The guide's grid has a limiting Kd about 0.33; above it the neighbour overtakes the centre, and beyond 0.5 the centre becomes negative. For a conservative spike, variance = 2Kd·t.

https://www.asksia.ai/cheatsheet/manchester-eart60061
Forward this to your tutorial group · Shared by ________
AskSiav2 · 2026-10-19 · side 1/2
Online open-book test (2 × 1.5 hours) after teaching week 12, Semester 1 2026/27 (check the date in your exam timetable).
Check your unit's exam rules for any limit on what may be brought in.
EART60061
Worked Steps · Traps · Fast Recall
Measuring and Predicting 1 · The University of Manchester
EART60061 cheatsheet QR code – open the full EART60061 study guide
Full guide + ask Sia
about EART60061
SIDE 2/2 WORKED EXAMPLES · COMMON ERRORS · METHOD CHOOSEROPEN-BOOK EXAM SHEETKeep calculations + exercise files organised

11 · Worked: Sample SpreadCh 2 example

Data: conductivity = 212, 220, 205, 218, 215 µS/cm.

205212x̄=214215218220δ: −9, −2, +1, +4, +6
All deviations are measured from 214; signed deviations cancel, so use absolute values or squares.
  1. x̄ = 1070/5 = 214 µS/cm.
  2. Deviations: −2, 6, −9, 4, 1; Σ=0.
  3. MAD=(2+6+9+4+1)/5=4.4 µS/cm.
  4. Squares: 4+36+81+16+1=138.
  5. s=√(138/4)=√34.5=5.87 µS/cm.

Trap: divisor is n−1=4 for a sample. The largest deviation (−9) receives extra weight after squaring.

12 · Drill: Normal TailCh 3 example

River temperature: x̄=14.2°C, s=1.8°C.

95%: 14.2 ± 1.96(1.8)=14.2 ± 3.53
10.67 to 17.73°C

Threshold 18.0°C: d=(18.0−14.2)/1.8=2.11. Upper-tail area ≈0.017 → 1.7% expected above 18.0°C.

PromptArea
within ±dcentral area
above +done upper tail
outside ±dboth tails
Say it This estimate assumes normal random variation and that the sample mean and s describe the population.

Fast drill: seven-foot logic

The guide's practice check uses x̄=171 cm, s=6.2 cm and threshold 185 cm. d=(185−171)/6.2=2.26; the upper-tail area is about 0.012, so roughly 1.2% exceed 185 cm. In Excel: 1−NORMDIST(185,171,6.2,TRUE).

Sanity checkExpected
Σ(xᵢ−x̄)0 (rounding aside)
Σ frequency classesn
normal-curve total area1
larger s, same nwider/lower curve
larger threshold dsmaller tail

Resolution: a coarse rectangle sum can give about 73% instead of 68%; refine class width. Density is not class frequency.

12b · Drill: Aim → ObjectivesCh 1 example

Bypass → swimming lake. General introduction: chloride harm → de-icing source → drainage path → recreational use.

13 · Worked: Two 95% Meanscourse example

Seaweed copper (mg/kg): P = 14.2,15.8,13.9,16.1; Q = 17.6,18.9,17.1,18.4. n=4, df=3, t=3.182.

131619P 15.00±1.77Q 18.00±1.28overlap
The intervals overlap only from 16.72 to 16.77—but overlap is still overlap.
PQ
x̄60/4=15.0072/4=18.00
s1.1110.804
SE=s/20.5550.402
CL=t·SE1.771.28
interval13.23–16.7716.72–19.28

Decision: intervals overlap slightly; a difference cannot be claimed at 95% confidence. Do not judge by means alone.

13b · Confidence TrapsCh 4

  • Do not use SD as the error of a mean.
  • Do not use 1.96 when small n requires t.
  • df for a mean is n−1; regression uses n−2.
  • TINV first argument is outside probability 0.05.
  • Mean and limit: same units and decimal places.
  • A bar without stated confidence is uninterpretable.

Precision grows as √n: quadrupling n halves SE, all else equal.

Threshold check: lead guideline

Six soil replicates: x̄=31.4 mg/kg, s=2.6 mg/kg; t=2.571 for df=5. SE=2.6/√6=1.061; CL=2.571×1.061=2.73 mg/kg. Interval 28.67–34.13 lies wholly below 35, so the mean is below the guideline at 95% confidence.

Graph choiceReason
scatter by distanceshows spatial relation required by objective
95% error barsmake uncertainty/comparison visible
reference/consent lineturns plot into a decision
points unconnectedno claim between sampled sites
legend belowstates bars and units unambiguously

Opposing corrections: replacing s with SE shortens the bar because means vary less than readings; replacing 1.96 with small-sample t lengthens it because few replicates give less confidence. The correct bar needs both.

14 · Worked: Nickel LoadCh 5 example

Given: Q=2.40±0.12 m³/s; C=0.050±0.006 mg/L.

2.40m³/sfe .0500.050mg/Lfe .1200.120g/sfe .13010.37kg/day±1.35concentration error dominates
Separate the number-and-unit chain from the fractional-error chain.
  1. Flux=2.40×0.050=0.120 g/s.
  2. Fractional errors: 0.12/2.40=0.050; 0.006/0.050=0.120.
  3. Combined fe=√(0.050²+0.120²)=√0.0169=0.130.
  4. Daily=0.120×86,400=10,368 g=10.37 kg/day.
  5. E=0.130×10.37=1.35 kg/day.

Answer: 10.37±1.35 kg/day (equivalently 0.120±0.016 g/s). Improve concentration measurement first because its fractional error is larger.

14b · Rule-and-Unit TrapsCh 5

MistakeCorrection
sum errors for A plus Bquadrature of absolute E
product errors in unitsquadrature of fractional E
give 86,400 an errorexact; fe unchanged
judge consent by centrecompare full interval
negative tidal flux = wrongcan mean net landward flux

Sampling logic: cover predictable tidal cycles at spring and neap; acknowledge unsampled storm variation and storage/resuspension.

Write every propagation answer in two passes

  1. List each measured quantity as value, units, absolute error and fractional error.
  2. Calculate the central result with units; do not mix error arithmetic into this line.
  3. Choose absolute quadrature for addition/subtraction or fractional quadrature for multiplication/division.
  4. Convert the final fractional error back to an absolute error in the answer's units.
  5. Identify the dominant contribution and compare the whole interval with the criterion.
OperationCarryQuick check
A±Babsolute Eanswer E exceeds each component E
A×B or A/Bfractional Eanswer fe exceeds each component fe
xⁿ|n| times fenegative n does not make negative error
× exact constantsame feabsolute E scales with result

15 · Drill: Regression SignificanceCh 7 example

Given: n=15 paired samples, r²=0.30.

r²=.30r=.548t=2.36p=.035df=n−2=13 · two-tailed
r² explains variation; p tests evidence. Report both.
  1. r=√0.30=0.548; df=15−2=13.
  2. t=0.548√13/√0.70 = 0.548×3.606/0.837 = 2.36.
  3. TDIST(2.36,13,2) gives p=0.035.
  4. p<0.05: significant at 95%, although discharge explains only 30% of sulphate variation.

High r² need not be significant when n is small: the guide's four-point check has r²=0.85 but p=0.078.

15b · Transform / Split / RecheckCh 7 case

Pattern / lawPlot or action
y=axᵇlog y vs log x
y=aeᵇˣln y vs x
time loopjoin points in time; split limbs
two solute sourcesseparate process contributions

A straight concentration–discharge line crossing zero is physically impossible under pure dilution. Use the process to choose the law. Splitting limbs captures depletion/history; each smaller subset has new df and needs its own p.

Case sequence: time series → power-law fit → time-ordered loop → split limbs → remove H⁺ contribution → refit and recompute p. Name the physical assumption relaxed at each change.

Different limbs at the same Q indicate solute depletion: hysteresis, not random scatter. A useful rating curve may predict while also revealing sources and storage.

16 · Drill: Logistic PredictionCh 8 example

Data: N₀=8, N₄=20, plateau K≈400.

r=(ln20−ln8)/4=0.916/4=0.229 month⁻¹
N₁₀=400·8/[8+392e⁻²·²⁹]=3200/47.7≈67
exponential: 8e²·²⁹≈79

Logistic is lower because resources already moderate growth. Estimate r only from the early straight segment of lnN vs t; then calibrate r and K, and validate against observations.

Limit / plotMeaning
N≪Knearly exponential
N=Kgrowth stops
early lnN straightgradient=r
plateauestimate K

17 · Worked: Two CompartmentsCh 9 example

F₀₁=500, F₀₂=200 units/y; K₁₀=.08, K₁₂=.02, K₂₀=.005 y⁻¹; X₁=3000, X₂=40,000; Δt=1 y.

X₁ 3000→3200X₁*=5000X₂ 40000→40060X₂*=60000+500+200−240−200+60
Year-one fluxes use starting stores: X₁→X₂ is 0.02×3000=60.
  1. X₁′=3000+500−(.08+.02)3000=3200.
  2. X₂′=40000+200+.02(3000)−.005(40000)=40060.
  3. X₁*=500/.10=5000.
  4. X₂*=[200+.02(5000)]/.005=60,000.

Both stores are below equilibrium and increasing: soil is a sink. Residence times: X₁=1/.10=10 y; X₂=1/.005=200 y. Never update X₁ then reuse the new value in X₂ during the same explicit step.

17b · Verify Before TrustingCh 9

CheckEvidence
coefficientevery ΣKout·Δt<1
time stephalve Δt; difference immaterial
boundarylong-run output = analytical equilibrium
balanceinputs−outputs explain Δstore

The course demonstration found one-year vs half-year steps differed by under 1%; smaller steps brought no useful benefit relative to input accuracy. State both the difference and why it is acceptable.

18 · Model-Term DecoderCh 6

One mixed reach = abstraction. Heat gain ∝ sunshine = assumption. Heat-balance equation = formulation. Hourly spreadsheet = numerical solution. Half-hour rerun = verification. Tune shading factor = calibration. New-period comparison = validation.

Term pairDo not confuse
conceptual / mathematicalprocess description / equations
analytical / numericalexact integrated form / stepped approximation
verification / validationsolve equations right / equations represent reality
calibration / validationtune plausible values / test resulting model
deterministic / probabilisticone fixed answer / distribution from parameter ranges

Start simple. Failure may expose sensitivity or a missing process; an engineering model must predict for its purpose.

19 · Worked: Diffusion Two StepsCh 10 example

Unit mass starts in the centre; Kd=0.2.

t0t1t21.00.20.60.20.04.24.44.24.04
Explicit diffusion spreads symmetrically while the cell total stays exactly one.
  1. t1: centre=1−2(.2)=.60; neighbours=.20.
  2. t2 centre=.6+.2(.2−1.2+.2)=.44.
  3. t2 neighbours=.2+.2(.6−.4+0)=.24; outer=.2(.2)=.04.
  4. Mass=.44+2(.24)+2(.04)=1.00.
  5. Variance=2(.24)(1²)+2(.04)(2²)=.80=2Kd·t.

The centre remains highest because Kd<⅓. A drifting total means a flux was omitted or counted wrongly. A degrading pollutant would deliberately lose mass; a conservative one must not.

19b · Stability PictureCh 10 figure

.33 equal.5 zerocentreneighbour
At Kd≈0.33 neighbour overtakes centre; beyond 0.5 the centre becomes negative.

Fᵢ→ᵢ₊₁=Kd(Cᵢ−Cᵢ₊₁). Update every cell from the same old row. Negative or alternating concentrations are numerical artefacts.

Random walk: 500 particles move one unit left/right per step; the frequency becomes normal and variance grows with iteration. Particle count must stay fixed.

CheckPass
symmetryequal neighbours
massΣC constant
spreadvariance=2Kd·t
stabilityKd below limit

20 · Final Method Chooserall chapters

ClueWrite first
replicate spreads with n−1
rare thresholdd=(x−x̄)/s; tail
uncertain meant·s/√n
A±E plus B±Eabsolute quadrature
A±E times B±Efractional quadrature
line significancer²→r→t, df n−2→p
https://www.asksia.ai/cheatsheet/manchester-eart60061
Forward this to your tutorial group · Shared by ________
AskSiav2 · 2026-10-19 · side 2/2
Online open-book test (2 × 1.5 hours) after teaching week 12, Semester 1 2026/27 (check the date in your exam timetable).
Online test · 100% · multiple-choice + short-answer

Want one for YOUR exact syllabus?

Sia is your free desktop study agent. Drop your The University of Manchester EART60061 slides — Sia builds a sheet tailored to YOUR exam. Better than this library because it knows YOUR materials.

↓ Download Sia · Free