Utafiti huu unachunguza ni kwa kiwango gani mabadiliko ya uso na ujazo wa dampo linalofanya kazi la taka ngumu za manispaa yanaweza kufuatiliwa kwa kuaminika kupitia Airborne Laser Scanning (ALS/LiDAR). Swali kuu la utafiti si tu “Je, ujazo unaweza kuhesabiwa kwa LiDAR?”. Watafiti pia wanatathmini Digital Terrain Model (DTM) inapaswa kutayarishwa kwa spatial resolution gani, jinsi horizontal-position errors zinavyogeuka kuwa vertical error katika terrain yenye mteremko, na kwa kiwango gani vegetation inaweza kuleta error inayoweza kuchanganywa na real ground-surface change.

Utafiti ulifanywa katika Strážnice municipal solid waste landfill katika eneo la Moravia nchini Czechia. Eneo la karibu 4 hectares limekuwa likitumika tangu 1993; sehemu moja bado inapanuliwa huku sehemu nyingine ikiwa imefungwa na kurehabilitiwa. Kwa hiyo, eneo hilo linajumuisha active waste-deposition zones, roads, sloping surfaces, herbaceous vegetation na shrubs ndani ya study area moja.

Airborne laser-scanning data zilikusanywa tarehe 28 May 2023 na 5 October 2023, kwa interval ya karibu miezi mitano. Katika flights zote mbili, average point density ilikuwa karibu 10 points/m². RIEGL LMS Q780 full-waveform laser scanner ilitumika kwenye Cessna 208B Grand Caravan photogrammetric aircraft katika flight altitude ya karibu 1 km.

Kutoka original point clouds, DTM yenye resolution ya 0,3 m ilitengenezwa; kisha data density ilipunguzwa na additional models zikatayarishwa kwa resolutions za 0,5 m, 0,8 m na 1,5 m. Hivyo athari ya DTM cell size kwenye calculated volume change ililinganishwa.

Kulingana na results, hakukuwa na statistically significant difference katika volume change kati ya 0,3 m na 0,5 m resolution (\(p=0,91\)). Vivyo hivyo, difference kati ya 0,3 m na 0,8 m haikuwa significant (\(p=0,64\)). Kwa upande mwingine, comparison ya 0,3 m na 1,5 m ilikuwa statistically significant (\(p=0,03\)). Kwa hiyo, watafiti wanapendekeza kuepuka 1,5 m resolution katika precise landfill-monitoring studies.

Scientific result inahitaji kutenganisha recommendations mbili tofauti za resolution. Range ya 0,3–0,8 m ni group ya resolutions iliyotoa reliable results katika tests; lakini data volume, detail na operational cost vinapotathminiwa pamoja, watafiti wanapendekeza range ya 0,3–0,5 m kama practical optimum balance.

Katika utafiti, average horizontal ALS-position error ilihesabiwa kuwa karibu 0,38 m. Katika flat ground, error hii inaweza kutosababisha large height error moja kwa moja; lakini kadiri slope inavyoongezeka, horizontal shift ileile hugeuka kuwa significant vertical error. Kwa kutumia average XY error ya 0,38 m, approximate vertical error ilihesabiwa kuwa 0,03 m kwa slope ya 5°, 0,10 m kwa 15°, 0,17 m kwa 25° na 0,23 m kwa 35°.

Vegetation ni separate error source kutoka resolution na katika baadhi ya cases inaweza kuwa kubwa zaidi. Katika comparison na independent GPS points, kwa data za 28 May, GPS–DTM mean difference katika 0,3 m DTM ilikuwa −0,20 m na standard deviation 0,21 m, huku kwa data za 5 October katika resolution hiyo hiyo values za −0,24 m na 0,33 m zikipatikana. Kwa 1,5 m resolution, October standard deviation iliongezeka hadi 0,42 m.

Kwa hiyo, watafiti wanaonyesha kwamba kutengeneza pixel ndogo pekee haitoshi. Katika LiDAR data, vegetation kuainishwa kimakosa kama ground, laser pulses kufika ardhini kwa proportions tofauti kupitia grasses na shrubs, na seasonal vegetation change vinaweza kuunda false volume changes zinazoonekana kama real landfill deformation.

Ili kupunguza effect hii, VNIR hyperspectral images katika range ya 0,4–1,1 µm zilikusanywa wakati mmoja na ALS data, green vegetation ikatambuliwa kwa NDVI, na herbaceous plants zikatenganishwa na tree/shrub areas kwa kutumia nDSM. Katika uncertainty model, height error ya 0,3 m ilipewa herbaceous vegetation na 0,4 m kwa trees na shrubs.

Field map ya utafiti haitoi tu volume change kwa location ileile, bali pia uncertainty ya change hii katika pixel level. Hivyo, badala ya kusema tu “kulikuwa na 0,2 m³ change hapa”, inatathminiwa pia ni kwa kiwango gani change hiyo inaaminika ndani ya error bounds za measurement system.

Kwa mtazamo wa Uturuki: Approach hii inatoa method inayoweza kutumika kufuatilia capacity use, filling progress, slope changes, settlement na rehabilitation areas katika sanitary landfills nchini Uturuki. Hasa katika large sites, LiDAR scans zinazofunika surface yote zinaweza kuonyesha spatial distribution ya volume na surface changes badala ya repeated classical point measurements. Hata hivyo, study imethibitishwa katika site moja tu nchini Czechia. Katika kuhamisha kwa facility nchini Uturuki, slope geometry, vegetation type, climate, flight altitude, sensor density, coordinate infrastructure na local GNSS control points zinapaswa kutathminiwa upya. Hakuna single na fixed DTM resolution kwa landfills zote nchini Uturuki inayoweza kutolewa kutoka study hii.

Tatizo kuu la traditional point measurement ni lipi?

Geodetic au GNSS-based field measurements zinaweza kutoa high accuracy katika specific points. Lakini landfill inapokuwa complex surface ya hectares kadhaa, kupima surface yote mara kwa mara kwa dense point intervals kunahitaji time, personnel na cost.

Tatizo muhimu zaidi ni kwamba point measurements hazionyeshi moja kwa moja local changes kati ya points. Waste accumulation katika sehemu moja ya landfill, settlement katika sehemu nyingine, na slope-geometry change mahali pengine vinaweza kutokea katika time interval ileile.

ALS inajaribu kutatua tatizo hili kwa kusample surface yote kwa millions of three-dimensional measurement points.

ALS na LiDAR ni kitu kilekile?

LiDAR ni remote-sensing method inayokokotoa distance kutoka travel time ya laser pulses kwenda na kurudi kutoka target surface. Airborne Laser Scanning (ALS) ni application form ambapo LiDAR sensor inatumika kwenye aircraft au other airborne platform.

Katika study hii, sensor haikuwa kwenye drone bali kwenye photogrammetric aircraft.

Study area

Test site ni Strážnice municipal solid waste landfill katika Moravia region nchini Czechia.

ALS campaigns mbili zilifanywaje?

Flight trajectories zilikokotolewa kwa GNSS na IMU data, deviations kati ya flight strips zikarekebishwa kwa corresponding planes na least-squares approach, kisha point clouds zikageoreferenced upya.

DTM inawakilisha nini?

Digital Terrain Model inawakilisha kadiri iwezekanavyo elevation surface ya bare ground. Katika landfill, tatizo ni kwamba si kila point inayoonekana na laser ni real ground.

Grass, shrub, tree, temporary object au irregularities kwenye waste surface vinaweza kuathiri ground classification. Kwa hiyo, DTM accuracy haitegemei technical accuracy ya laser sensor pekee.

Kwa nini DTM resolutions nne zilijaribiwa?

0,3 m DTM iliyotengenezwa kutoka initial density ya karibu 10 points/m² ndiyo basic high-resolution surface.

Point cloud kisha ilithinned ili kutengeneza:

models.

Lengo si kuchukulia smallest pixel automatically kama “best”, bali kubaini coarse resolution inaanza lini kusababisha meaningful information loss katika volume-change estimation.

Kwa nini random seed ilijaribiwa?

Point cloud ilipothinned, points ndani ya kila grid cell zilichaguliwa randomly. Kwa kuwa random selection yenyewe inaweza kuleta new uncertainty, watafiti walitumia four different seeds kama 100, 200, 300 na 400.

Method hii hufanya process iwe reproducible kwa kuhakikisha point subset ileile inachaguliwa tena kwa seed ileile.

Jumla ya 24 DTM zilitengenezwa kutoka thinned point clouds na ikatathminiwa kama different seed combinations zilibadilisha height results significantly.

Seed result

Statistical tests zilionyesha kwamba seed selection haikuunda significant systematic effect kwenye DTM elevations. Kwa mfano, kwa 0,5 m resolution, p-values kwa three control points zilikuwa 0,99, 1,00 na 0,98 mtawalia.

Kwa 0,8 m resolution, values zilikuwa 0,86–0,96, huku kwa 1,5 m resolution source ikitoa range ya 0,48–0,99. Values hizi zote ziko juu ya 0,05 significance threshold.

Kwa hiyo, conclusion ni kwamba random thinning process iliyotumiwa haikuzalisha systematic bias chini ya conditions zilizochunguzwa.

Angalizo kuhusu p-value expression katika source text

Katika results section ya source, expression “all p-values > 0.86” inatumika kwanza, na mara baada yake p-values za 0,48–0,99 zinaripotiwa kwa 1,5 m resolution.

Value ya 0,48 bado ni statistically nonsignificant kulingana na threshold ya 0,05; hivyo haibadilishi main conclusion ya study kuhusu seed. Hata hivyo, statement “all values are greater than 0,86” haiendani na values zilizoripotiwa na source yenyewe.

Je, resolution inabadilisha volume calculation kweli?

Paired t-test ilifanywa kwa kutumia 0,3 m resolution kama reference:

Result hii inaonyesha kwamba 1,5 m cell size inaweza kugeneralize surface kupita kiasi na kuleta significant variability katika volume-change estimate.

0,3–0,8 m au 0,3–0,5 m?

Statements mbili katika sections tofauti za source hazipingani kabisa.

Kwa upande wa statistical result, 0,3; 0,5 na 0,8 m ni tested reliable resolutions kwa sababu hakuna significant volume difference iliyopatikana kati ya 0,3 m na 0,8 m.

Kwa operational recommendation, watafiti wanazingatia pia data density na processing burden na kutathmini 0,3–0,5 m range kama balance inayofaa zaidi.

Kwa nini horizontal-position error hugeuka kuwa vertical error kwenye sloping site?

Ikiwa LiDAR point imesogea kidogo kutoka actual horizontal location yake, height difference inaweza kuwa limited katika flat ground. Horizontal shift ileile ikitokea kwenye steep slope, point huanguka kwenye sehemu ya slope iliyo juu au chini zaidi.

Study ilitumia simplified relation:

\[ \varepsilon_Z = \varepsilon_{XY}\tan(S) \]

ilitumika kwa namna hii.

Hapa:

• \(\varepsilon_Z\): vertical height error,
• \(\varepsilon_{XY}\): horizontal position error,
• \(S\): terrain slope.

Measured XY error level

Katika comparisons kwenye three fixed, horizontal na asphalt control points:

• mean horizontal-error range: 0,24–0,54 m,
• SD range: 0,20–0,34 m,
• RMSE range: 0,31–0,62 m,
• mean \(\varepsilon_{XY}\): 0,38 m,
• mean SD: 0,28 m,
• mean RMSE: 0,48 m

ziliripotiwa.

Watafiti wanaonya kwamba changes ndogo kuliko karibu 0,20–0,34 m zinaweza kuwa difficult kutenganishwa reliably na random positional noise katika dataset iliyotumiwa.

Error kadiri slope inavyoongezeka

Kwa hiyo, apparent height changes za centimeters chache kwenye steep landfill slopes hazipaswi kutafsiriwa moja kwa moja kama settlement au filling bila kuzingatia effect ya horizontal georeferencing error.

Vegetation ilitambuliwaje?

Katika dates zilezile za ALS, VNIR data zilikusanywa katika spectral range ya 0,4–1,1 µm kwa kutumia CASI-1500 imaging spectrometer. Spatial resolution ya VNIR images ni 0,8 m.

NDVI ilitumika kutenganisha green vegetation na pixels zenye:

\[ NDVI>0,4 \]

zikaainishwa kama vegetated.

Objects zilizo 0,7 m juu ya ground level kwenye nDSM zilitumika kutenganisha trees, most shrubs na similar tall structures.

Hivyo, uncertainties tofauti ziliweza kufafanuliwa kwa:

• herbaceous vegetation chini ya 0,7 m,
• tree na shrub areas juu ya 0,7 m

.

Vegetation-derived error values

Katika uncertainty model ya study:

• herbaceous vegetation: 0,3 m height error,
• trees na shrubs: 0,4 m height error

ziliwekwa.

Values hizi hazikuchaguliwa kwa kutegemea nominal LiDAR accuracy ya machine manufacturer pekee, bali independent GPS comparisons kwenye site na vegetation-derived DTM errors katika literature pia zilizingatiwa.

Volume change ilihesabiwaje?

Height-change raster kati ya dates mbili iliundwa kama:

\[ DTM_{changes} = DTM_{t1}-DTM_{t2} \]

iliundwa kwa namna hii.

Height change katika kila pixel huzidishwa na pixel area ili kuibadilisha kuwa volume change:

\[ Volume_{changes} = DTM_{changes}\times A_{pixel} \]

Kwa 0,3 m DTM, pixel area ni:

\[ 0,3\times0,3 = 0,09\ \mathrm{m^2} \]

.

Uncertainty iliunganishwaje?

Kwa DTM moja, two main error components zilichukuliwa pamoja:

\[ \sigma_{total} = \sqrt{ \sigma_1^2+\sigma_2^2 } \]

Hapa \(\sigma_1\) ni vertical error inayotokana na XY-position error kwenye slope, na \(\sigma_2\) ni error component inayotokana na vegetation.

Surfaces za dates mbili tofauti zinapolinganishwa, uncertainty ya DTM zote mbili huzingatiwa:

\[ \sigma_{changes} = \sqrt{ \sigma_{total,1}^2+ \sigma_{total,2}^2 } \]

Approach hii huruhusu kutengeneza si difference map pekee bali pia reliability map ya difference.

Ramani kwenye page 7 inaonyesha nini?

Katika Figure 4 ya source, kuna raster maps mbili side by side kwa landfill ileile. Ramani ya kushoto inaonyesha volume change kati ya May–October, huku ramani ya kulia ikionyesha calculated uncertainty ya change hiyo hiyo.

Interpretation ya maps hizi mbili pamoja ndiyo main advantage ya method. Surface change inayoonekana kubwa haipaswi kutathminiwa moja kwa moja kama physical deformation ikiwa uncertainty pia ni kubwa katika region ileile.

Figure 5 inaonyesha operational workflow kama GPS reference measurement + two-period ALS + VNIR → DTM/nDSM → position, slope na vegetation errors → height change → volume change na uncertainty.

Tofauti kati ya active waste-deposition zones na stable surfaces

Kulingana na discussion ya source, katika waste_1 area, mean height change ni karibu 0,3–0,4 m across resolutions na standard deviation ni karibu 0,02 m.

Katika active area iliyoitwa waste_2, height change ni karibu 5,2–5,5 m na SD ni karibu 0,07 m. Magnitude hii inaonyesha large-scale surface elevation inayoendana na real waste-deposition/shaping activities badala ya data resolution au measurement noise.

Kwenye road surface, mean height change ni karibu 0,003 m na SD karibu 0,01 m, ambazo ni ndogo sana. Hii inaunga mkono kwamba road ilifanya kazi kama stable reference surface kati ya dates mbili.

Kwa nini vegetation inaweza kutoa false volume change?

LiDAR pulse inaweza kurudi kutoka upper au intermediate layers za dense herbaceous vegetation katika period moja, huku katika period nyingine ikipenya deeper kati ya dried plants na kufika ground.

Hata kama actual soil surface haijabadilika kabisa katika dates mbili, DTM inaweza kubadilika kama:

May: measurement karibu na upper vegetation surface → October: deeper penetration hadi ground

inaweza kubadilika kwa namna hii.

Difference raster inaweza kuonyesha hili kimakosa kama negative ground change au compaction.

Study inahusisha baadhi ya volume decreases zilizoonekana katika central region na seasonal vegetation change badala ya actual waste settlement.

ALS data inapaswa kukusanywa katika season gani?

Watafiti wanapendekeza kwamba ili kupunguza vegetation effect, ALS campaigns zifanywe inapowezekana katika early spring au late autumn, wakati vegetation cover iko minimum.

Lengo hapa si tu kufanya laser pulses nyingi zaidi zifikie ground, bali pia kupunguza kuchanganya difference ya plant phenology kati ya two measurement periods na real terrain change.

Mbinu na Matokeo ya Utafiti

Data sources

GPS reference structure

Kwa co-registration ya ALS datasets, three GPS control points zilitumika katika flat, asphalt na assumed-unchanged areas nje ya landfill.

Kwa independent validation ya DTM elevation kwenye actual field, 12 additional GPS reference points ziliwekwa katika areas zilizoathiriwa na terrain irregularity na vegetation.

Chanzo kinaeleza wazi kwamba number of points ilikuwa limited kutokana na access restrictions.

GPS–DTM accuracy results

Main result inayoonyeshwa na Table 3

Katika dates zote mbili, mean absolute GPS–DTM difference kwa ujumla huongezeka kadiri resolution inavyokuwa coarse.

Hasa katika October data, standard deviation:

\[ 0,33\rightarrow0,37\rightarrow0,39\rightarrow0,42\ \mathrm{m} \]

huongezeka kwa namna hii.

Hali hii inaonyesha kwamba coarser DTM zinaweza kuongeza systematic error chini ya surface generalization na vegetation effects.

Kwa nini accuracy limit katika abstract inapaswa kusomwa kwa uangalifu?

Abstract ya source inasema kwamba 0,3 na 0,5 m resolutions zina “mean difference ≤0,23 m, SD ≤0,24 m”.

Limits hizi zinaendana na 28 May Table 3 results:

• 0,3 m: −0,20 ± 0,21 m
• 0,5 m: −0,23 ± 0,24 m

Hata hivyo, katika 5 October results:

• 0,3 m: −0,24 ± 0,33 m
• 0,5 m: −0,27 ± 0,37 m

zimetolewa.

Kwa hiyo, ≤0,23 m na ≤0,24 m limits katika abstract hazipaswi kutafsiriwa kama overall maximum values za campaigns zote mbili.

Source-internal note kuhusu SD statement katika vegetated areas

Source abstract inasema kwamba katika vegetated areas, SD “increased by up to 0,10 m” kadiri DTM resolution ilivyobadilika.

Katika Table 2, SD values za herbaceous area ni:

\[ 0,04,\ 0,04,\ 0,06,\ 0,02\ \mathrm{m} \]

na SD values za shrub area ni:

\[ 0,15,\ 0,15,\ 0,05,\ 0,14\ \mathrm{m} \]

.

Largest absolute resolution difference katika shrub area kwa kweli ni 0,10 m; lakini direction ni decrease kutoka 0,15 m katika 0,3 m hadi 0,05 m katika 0,8 m. Kwa hiyo, “increase” statement katika source abstract haiendani kikamilifu na direct values za Table 2.

Dependence ya volume-change results kwa surface type

Highest uncertainty iko wapi?

Source inasema vegetation inaweza kuwa na stronger effect kwenye DTM accuracy kuliko raster resolution.

Katika tree na shrub areas, volume-change uncertainty ilifikia karibu 0,05 m³/pixel kwenye map.

Kwa hiyo, high-resolution LiDAR peke yake si guarantee ya high accuracy katika vegetated surfaces.

Matokeo yanayoungwa mkono na utafiti

• ALS ni applicable method kwa repeated surface na volume-change analysis katika operational municipal solid waste landfill.
• Hakuna statistically significant difference katika volume change kati ya 0,3 m na 0,5 m DTM results.
• Difference kati ya 0,3 m na 0,8 m pia si significant.
• Difference kati ya 0,3 m na 1,5 m ni significant na 1,5 m haionekani suitable kwa precise operational monitoring.
• 0,3–0,5 m range inapendekezwa kama practical optimum kwa data quantity na accuracy.
• Horizontal-position error inaweza kugeuka kuwa significant vertical height error kadiri slope angle inavyoongezeka.
• Vegetation inaweza kuwa stronger error source kuliko DTM resolution.
• VNIR/NDVI na nDSM layers zinaweza kusaidia spatial modeling ya vegetation-derived uncertainty.
• Kutengeneza uncertainty raster pamoja na volume change kunaimarisha separation ya real change na measurement uncertainty.
• ALS campaigns zinapendekezwa kufanywa early spring au late autumn wakati vegetation cover ni ndogo.

Matokeo ambayo utafiti haujathibitisha

• Haijaonyeshwa kwamba 0,3–0,5 m DTM resolution ni universal optimum kwa landfills zote duniani.
• Study ilifanywa katika landfill moja tu nchini Czechia.
• Independent field validation ilifanywa kwa 12 GPS points tu ndani ya landfill.
• Study haipimi methane emissions, leachate au groundwater contamination moja kwa moja.
• Haijaonyeshwa kwamba volume decrease yote inaweza kutafsiriwa kama actual waste settlement; baadhi ya changes zinaweza kutokana na vegetation.
• Haijaonyeshwa kwamba ALS huondoa kabisa functions zote za classical geodetic measurement; GPS control points ni sehemu muhimu ya method.
• Haijaonyeshwa kwamba seasonal vegetation effect imeondolewa kabisa.
• Haijathibitishwa kwamba error magnitudes zilezile zitapatikana kwa different LiDAR sensors, different flight altitudes na different climates.

Main limitations

Limitation ya kwanza ni kwamba study inategemea landfill moja. Ingawa watafiti wanasema method inaweza kutumika katika other sites, wanakubali kwamba additional validation inahitajika chini ya different landfill geometries na vegetation conditions.

Limitation ya pili ni kwamba kulikuwa na 12 independent GPS validation points tu kwenye site. Number kubwa zaidi na balanced distribution ya validation points inaweza kuongeza statistical power ya geometric-accuracy estimate.

Limitation ya tatu ni kuwepo kwa baadhi ya operator-dependent steps katika workflow. Selection na positioning ya GPS control points na interpretation ya ALS-filtering results zinaweza kuleta human-originated bias.

Limitation ya nne ni kwamba ALS campaigns mbili zilifanywa chini ya different vegetation-phenology conditions. Ingawa hii pia iliruhusu study kuonyesha vegetation effect, inafanya separation ya real surface change kuwa difficult zaidi.

Operational application recommendation

General workflow iliyoundwa katika source imegawanywa katika sehemu mbili kuu:

Data collection:

• measurement ya GPS reference points,
• ALS scanning katika dates mbili au zaidi,
• collection ya VNIR data kwa vegetation-derived uncertainty.

Data processing:

• ALS co-registration,
• DTM na nDSM creation,
• calculation ya XY- na slope-derived Z error,
• NDVI-based vegetation mask,
• height change,
• uncertainty propagation,
• volume-change na volume-uncertainty maps.

Watafiti wanaeleza kwamba aircraft-based ALS inaweza kuwa advantageous wakati multiple large landfills zinahitaji kuscanned ndani ya muda mfupi; lakini kwa operational monitoring ya single site, UAV-based laser scanning inaweza kuwa suitable zaidi.

Maelezo ya Chanzo na Mbinu

Jina kamili asilia la utafiti: Detecting volume changes in municipal solid waste landfill using airborne laser scanning

Waandishi: O. Brovkina; M. Pikl; F. Zemek; J. Michálek.

Corresponding author: O. Brovkina.

Taasisi: Global Change Research Institute, Czech Academy of Sciences, Department of Remote Sensing, Brno, Czech Republic; Municipal Office Strážnice, Czech Republic; University of South Bohemia in České Budějovice, Faculty of Agriculture and Technology, Czech Republic.

Aina ya chanzo: Peer-reviewed research article.

Jarida: Waste Management Bulletin.

Mchapishaji: Elsevier B.V.

Publication: Volume 4, Issue 1, 2026, Article 100272.

First publication online: 3 December 2025.

Issue date: April 2026.

DOI: 10.1016/j.wmb.2025.100272.

Official link:https://doi.org/10.1016/j.wmb.2025.100272

ISSN: 2949-7507.

Leseni: Creative Commons Attribution (CC BY 4.0).

Funding/support: Study iliungwa mkono na Technology Agency of the Czech Republic kupitia grant number SS06020164 katika Environment for Life program na grant number LM2023048 ya CzeCOS program chini ya Czech Ministry of Education, Youth and Sports.

Conflict of interest: Waandishi walieleza kwamba hakuna known financial interest au personal relationship inayoweza kuathiri study.

Data availability: Source text haina separate “Data Availability” section; kwa hiyo hakuna additional claim kwamba raw ALS, VNIR au GPS datasets zinapatikana publicly.

Author contributions

Kulingana na CRediT statement katika source, O. Brovkina alifanya original draft, project administration, methodology na investigation; M. Pikl review/editing, software, investigation na formal analysis; F. Zemek review/editing, visualization na methodology; J. Michálek visualization na validation.

Source-internal consistency note 1 — random-seed p-values

Katika results section, statement “all p-values > 0.86” inaonekana kwanza kwa different seed values. Paragraph hiyo hiyo inaripoti p-values za 0,48–0,99 kwa 1,5 m resolution.

Difference hii haibadilishi statistical conclusion ya study; 0,48 pia iko juu ya 0,05 significance threshold. Hata hivyo, statement “all values are greater than 0,86” haiendani na lowest value iliyotolewa katika source.

Source-internal consistency note 2 — fine-DTM accuracy limits

Abstract inasema mean difference ya 0,3 na 0,5 m DTMs ni ≤0,23 m na SD ni ≤0,24 m. Values hizi zinatoa summary sahihi ya 28 May Table 3 results.

Hata hivyo, 5 October Table 3 results ni −0,24 ± 0,33 m kwa 0,3 m na −0,27 ± 0,37 m kwa 0,5 m. Kwa hiyo, thresholds hizi si absolute upper bounds zinazofunika ALS campaigns zote mbili.

Source-internal consistency note 3 — vegetated-area SD change

Abstract inasema SD katika vegetated areas “increased” hadi 0,10 m katika coarser DTM resolutions. Katika Table 2, largest 0,10 m difference kwa shrub area hutokea kutoka 0,15 m hadi 0,05 m; yaani direction ni decrease.

Kwa hiyo, katika Verianla text, “0,10 m increase” haijarudiwa kama direct scientific result; explicit numerical values za Table 2 zimetolewa.

Evidence na generalizability limit

Research inatoa strong applied evidence kwa kutumia real operational site, two real ALS campaigns na independent GNSS measurements. Kwa upande mwingine, experimental design ni single-site na number ya validation points ni limited.

Kwa hiyo, recommended resolutions na error magnitudes hazipaswi kuhamishwa moja kwa moja kama fixed standard kwa sanitary landfills katika countries nyingine au different surface conditions.

Method, equations, measurement values, statistics, results na limitations katika Verianla explanation hii zinategemea source study iliyochunguzwa. Field data ambazo source haijaripoti hazijaongezwa; numerical inconsistencies ndani ya source hazijasahihishwa kimya kimya kwa kutumia external information au assumptions.