Wyniki wyszukiwania - BazTech

1

Selection of optimal tree top detection parameters in a context of effective forest management

Kolendo Łukasz, Ksepko Marek

Ekonomia i Środowisko

|

2019

|

nr 1

67--85

EN

In the process of tree stand parameter estimation based on data from airborne laser scanning ALS, the detection of a single tree is an important starting point. The aim of this work is to develop optimal values of parameters in the process of detection of tops and segmentation of stands on the basis of ALS data analysis. The research was carried out on the basis of ALS data from raids carried out in 2007 and 2017 on a fragment of the Zajma forest district in the Zednia forest inspectorate (north-eastern Poland). Parameters analyzed included: Ground Sampling Distance [m], the level of smoothing of the Canopy Height Model (CHM) with the Gaussian filter (the size of the moving window, the value of standard deviation), the filtration of the output point cloud, as well as the application of the additional interpolation algorithm CHM based on the analysis of raster cells neighborhood. The research has shown that it is possible to indicate detection parameters that ensure a very high correlation between the number of automatically detected treetops and the number of trunks found during fieldwork. Importantly, the optimal detection parameters developed for remote-sensing materials from the years 2007 and 2017 differ slightly, which ensures generally high accuracy of ALS data and the possibility of implementing the values of these parameters in other research objects.

2

The influence of number and size of sample plots on modelling growing stock volume based on Airborne Laser Scanning

Stereńczak K., Lisańczuk M., Parkitna K., Mitelsztedt K., Mroczek P., Miścicki S.

Drewno : prace naukowe, doniesienia, komunikaty

|

2018

|

vol. 61, nr 201

5--22

EN

Current forest growing stock inventory methods used in Poland are based on statistical methods using field measurements of trees on circular sample plots. Such measurements are carried out with traditional equipment, i.e. callipers and range finders. Nowadays, remote sensing based inventory techniques are becoming more popular and have already been applied in North America and some Scandinavian countries. Remote sensing based forest inventories require a certain amount of ground sample plots, which serve either as reference data used for model calibration and/or as a validation dataset for the assessment of the accuracy of modelled variables. Using a set of 900 ground sample plots and Airborne Laser Scanner (ALS) from the Milicz forest district, a statistical model for the estimation of plot growing stock volume was developed. Next, the developed model was once again fitted to different variants of sample plot size and number of sample plots. Each variant was selected from a full 900 sample plot set. The selection started from 800, 700, 600, …, down to 25 plots, respectively, and was carried out in proportion to the dominant tree age range. To account for the area effect, each plot number variant was similarly tested with various sample plot areas, i.e. 500, 400, …, 100 m2. Sampling in each variant was repeated in order to take into account the effect of a single selection. The results showed a strong relationship between obtained modelling errors and the size and number of used sample plots. It has been demonstrated that the number of sample plots has no influence on the accuracy of GSV estimation above about 300-400 sample plots (about 500 sample plots for bias), whereas sample plot size has a visible impact on estimation accuracy, which reduces with decreasing sample plot size, regardless of the number of sample plots. If it is about precision, results showed that the influence of a single selection to be relevant only below 300-400 plots (about 500 for bias) and the same trend can be observed in each sample plot size variant. The results showed it is possible to strongly reduce the number of ground sample plots (minimum 300- 400), while still maintaining decent accuracy and precision levels, at least in similarly investigated forest conditions.

3

Geodistributed analysis of forest phytomass: Subcarpathian voivodeship as a case study

Striamets O., Lyubinsky B., Charkovska N., Stryamets S., Bun R.

ECONTECHMOD : An International Quarterly Journal on Economics of Technology and Modelling Processes

|

2014

|

Vol. 3, No 1

95--104

EN

An approach to the implementation of spatial inventory of greenhouse gas sinks and emissions in forestry sector is presented. The algorithms for disaggregation of official statistical data on forests at the regional level, and formation of geodistributed database using the digital map of land use are proposed. As an example the forests of Subcarpathian Voivodeship of Poland were investigated. This study presents the results of modeling the flow of carbon in forest phytomass taking into account different tree species composition, age class and other characteristics. The ‘regional’ coefficients for phytomass and carbon deposited were clarified taking into account regional specificity. Also an algorithm for determining aboveground and underground phytomass of tree by its stock and forests of different types has been created. The correlation between the total area and stock for each species of forest-forming stands was analysed. The multilayer digital maps of deposited carbon, and greenhouse gas emissions in forests of Subcarpathian Voivodeship were created. For the spatial inventory of forest phytomass the territory of voivodeship was divided into square areas 2x2 kilometers in size, and stock phytomass in each such elementary area has been analised. Acheaved results were verified using GIS technology, and forest inventory data.

4

Estimating forest biomass by remote sensing radar data in Brazil

Santos dos J. R., Gama F. F., Bispo P. C.

Drewno : prace naukowe, doniesienia, komunikaty

|

2014

|

vol. 57, nr 192

119--132

EN

Remote sensing-radar was used to analyze forest mapping and biomass estimates on Brazilian territory . Two examples of SAR attributes for the modeling of the aboveground biomass of forest stands are presented: (1) full-polarimetric attributes of PALSAR/ALOS (Phased Array type L-band Synthetic Aperture Radar /Advanced Land Observing Satellite) for modeling in the Amazonian tropical forest, considering the influence of the geomorphometric aspects on this radar response, and (2) polarimetric and interferometric airborne data (XHH and full-polarimetric of P-band ) for modeling Eucalyptus sp. stands. In both cases, an analysis of forest structure variability through polarimetric signatures was conducted. A multivariate regression technique was used to integrate the variables from polarimetric and /or interferometric radar attributes and field inventory. Considering the terrain aspects where the tropical forest was located, the most significant variables for the biomass modeling were the Volumetric Scattering of Freeman-Durden target decomposition, Anisotropy, Relief Elevation, Slope, and the first and third helicity components of the Touzi model . For the Eucalyptus biomass model, the Interferometry Heigt and Canopy Scattering Index variables were significant. The statistical analysis based on field survey measures to validate each model, indicated a margin of error below 20% for the biomass estimations , showing the importance of SAR attributes for models of natural and planted forest stock density .

5

Estimates and forecasts of forest biomassand carbon sequestration in North America and Australia: a forty-five year quest

Botkin D. B., Ngugi M. R., Doley D.

Drewno : prace naukowe, doniesienia, komunikaty

|

2014

|

vol. 57, nr 192

7--28

EN

A half-century of forest inventory research involving statistically-valid field measurements (using statistically representative sample size and showing confidence limits) and well-validated forecasting methods are reviewed in this paper. Some current procedures overestimate global and large-scale forest biomass, carbon storage, and carbon sequestering rates because they are based on statistically-invalid methods (errors in estimates are unavailable and unreported), or they fail to consider key dynamic characteristics of forests. It is sometimes assumed that old-growth forests can serve as fixed, steady-state storage of biomass and carbon for indefinitely long periods, but it is shown by both modelling and remote sensing that forests are dynamic systems, the state of which can change considerably over as short a time as a decade. Forecasting methods show that maximum biomass and carbon storage in some important forest types occurs in mid-succession, not in old-growth. It is proposed, therefore, that realistic biomass and carbon storage estimates used for carbon credits and offsets be determined as the statistical mean minus the confidence interval and that practical carbon sequestering programs include specific timeframes, not indefinitely long periods of time.

6

Określanie parametru zagęszczenia drzew w drzewostanach sosnowych na podstawie analizy chmury punktów naziemnego skaningu laserowego

Wężyk P, Tompalski P.

Roczniki Geomatyki

|

2010

|

T. 8, z. 7

83-90

PL

Pomimo, iż zasięg skanera FARO LS 880 wynosił aż 76 m (Faro, 2007), to w warunkach analizowanego drzewostanu sosnowego, w strefie powyżej 19,54 m (analizowana powierzchnia ok. 1200 m2), różnica w określaniu zagęszczenia drzew przekroczyła 5% w stosunku do danych referencyjnych. Wynika to przede wszystkim z faktu wzajemnego przesłaniania się pni drzew oraz obecności podszytu, który również jest barierą dla promieni lasera. Należy podkreślić, iż parametr zagęszczenia drzew określany na większej powierzchni (> 500 m2) powinien lepiej reprezentować całe wydzielenie, jednak sposób rozmieszenia stanowisk skanera nie był dla tego celu zoptymalizowany. Pozycje skanera mogłyby być oddalone bardziej od siebie i należałoby zwiększyć liczbę sfer referencyjnych (np. do 10÷15 szt.), jak i rozdzielczość samego skanowania, którą obniżono do 1/4 możliwości modelu FARO LS880. Lokalizacje stanowisk skanera były dobierane jedynie pod kątem standardowych procedur związanych z inwentaryzacją lasu na kołowych powierzchniach próbnych (500 m2). Można sądzić, iż czynnikiem który miał decydujący wpływ na dokładność określenia liczby drzew (tym samym zagęszczenia) było rozmieszczenie stanowisk skanera w drzewostanie, których lokalizacja była determinowana przez rozmieszczenie pni drzew na powierzchni kołowej. Rozszerzenie analiz poza standardową kołową powierzchnię próbną wskazuje, iż można wykorzystać pozyskane dane TLS w celu zwiększenia próby, co zapewnia większa powierzchnia pozyskania danych. Im większa powierzchnia próbna tym lepiej odzwierciedla ona charakter drzewostanu, jednak parametr zagęszczenia nie powinien być określany powyżej progowej wartości (oko?o 19,5 m). Uzyskanie odpowiedzi na pytanie - czy skanować więcej mniejszych powierzchni czy też mniejszą liczbę większych obszarów - wciąż wydaje się niemożliwe bez odpowiednich danych referencyjnych i integracji wielu źródeł informacji, takich jak lotniczy skaning laserowy czy pomiar DGPS.

EN

Modern multifunctional forestry model requires accurate knowledge of the amount of wood stock, which is controlled by means of forest inventory. One of many important forest parameters is density understood as the number of trees per area unit (1ha). This paper presents a method for defining this parameter with the use of terrestrial laser scanning data (TLS), based on data from one or a couple of scans (circular plot). The analysis was performed on 5 plots located in 107 year old pine stand (Milicz Forest District), on which TLS data was collected with the use of FARO LS 880. The results are highly correlated with ground-truth data from forest inventory. Increasing the radius of the analysis outside the borders of study plot leads to lowering of the density parameter, which is mainly caused by the shadowing effect. The maximum measurement distance of the FARO scanner is equal to 76m but stems and understory are limiting this value. Over 20m (r=19,54m; 1200m2) the errors of the TLS method exceed 5%. Further work of the authors is directed at selecting statistically representative TLS plots based on Airborne Laser Scanning.

7

Aktualizacja baz danych SILP oraz leśnej mapy numerycznej w oparciu o dane z lotniczego skaningu laserowego

Wężyk P., Szostak M., Tompalski P.

Archiwum Fotogrametrii, Kartografii i Teledetekcji

|

2010

|

Vol. 21

437--446

PL

Celem pracy było zaimplementowanie danych pozyskanych technologią lotniczego skaningu laserowego (ALS) w zautomatyzowanej procedurze aktualizacji granic pododdziałów, tj. przebiegu wektorów obiektów podstawowych Leśnej Mapy Numerycznej (LMN) oraz weryfikacji atrybutu wysokości drzewostanów, zapisanej w bazie danych Systemu Informatycznego Lasów Państwowych (SILP). Obiektami testowymi były obręby w Nadleśnictwach Milicz (Milicz) i Chojna (Piasek). Obszar badań w każdym z obrębów obejmował ok. 6 000 ha lasów, z czego blisko 80% stanowiły drzewostany sosnowe. Prace rozpoczęto od aktualizacji przebiegu granic wektora (SHAPE) pododdziałów zapisanych w bazie geometrycznej LMN w oparciu o modele generowane z chmury punktów ALS oraz obraz cyfrowej ortofotomapy. Następnie przeprowadzono aktualizację atrybutowej bazy danych SILP/LAS. Określenie wysokości całego drzewostanu oparto na powierzchni pododdziału z wyłączeniem luk, wykorzystując analizę chmury punktów ALS, tj. 95 percentyl. W celu porównania wyników do danych referencyjnych zbieranych metodami tradycyjnymi (SILP), wygenerowano modele rastrowe (GRID) wysokości drzewostanów określone metodą ALS (HALS) oraz HSILP. W obu obrębach stwierdzono zaniżenie wartości wysokości z bazy SILP/LAS. Średnia różnica (HDiff) wyznaczenia wysokości metodą ALS w stosunku do SILP wyniosła dla obrębu Piasek i Milicz, przy uwzględnieniu znaków odchyłek, odpowiednio +0.9 m oraz +2.3 m, natomiast w przypadku wartości bezwzględnych 2.1 m oraz 3.2 m. Ustalono, że zasadniczą rolę w wartości błędu odegrały licznie występujące drzewostany młodszych klas wieku.

EN

Automatic processing of remotely sensed data, like ALS point clouds, is crucial for modern economy, including forestry. The aim of the study was to develop automated procedures for digital forest map (LMN) revision and automated verification of the attributes (height) stored in the forest descriptive database (SILP), both based on airborne laser scanner datasets. The study areas were the Piasek (Chojna) and Milicz management forest districts, covering about 6,000 ha (80% Scots pine stands). The workflow of verifying and updating a digital map started with updating the compartment borders, which was based on nDSM (created from classified point cloud) and digital ortophoto (RGB+NIR) as well. The developed method, based on normalized ALS point cloud and GIS analysis, provided instant possibility for compartment border update, revealing additional objects like gaps or tree biogroups. The total area of automatically detected objects was around 15% lower when compared to the reference data for Chojna forest district and 10% higher regarding Milicz forest district. Around 84.0% and 85.5% of the gaps matched the reference for Chojna and Milicz forest districts, respectively. A method based on point cloud distribution (95th percentile) within compartment borders to assess its height was presented in the study. The results were compared to a height model (GRID) generated from descriptive database. For both the study areas the height stored in SILP database was lower than the height value derived from ALS data. The difference was equal to +0.9 m (Chojna; absolute difference 2.1 m) and +2.3 m (Milicz; absolute difference 3.2 m). When the stand area was used as a weight in the difference calculation, the difference values (HDiff) changed to +0.6 m (Chojna; absolute difference1.5 m) and +2.4 m (Milicz; absolute difference 2.7 m). Concerning the deciduous stands, the difference was higher (~+1 m) than for the Scots pine stands. The analysis performed confirms the possibility of using airborne laser scanning for geometrical (LMN) and descriptive (SILP/height) database updating. Periodical stand monitoring based on ALS technology can guarantee keeping the databases up to date without the necessity of costly and time consuming field measurements.

8

Modelowanie chmury punktów ze skaningu laserowego w obszarze koron drzew

Wężyk P.

Archiwum Fotogrametrii, Kartografii i Teledetekcji

|

2008

|

Vol. 18b

685--695

PL

Celem pracy było określenie przydatności i zaproponowanie rozwiązań automatyzacji procesu określania wybranych parametrów taksacyjnych drzewostanów związanych z koronami drzew, w oparciu o dane z lotniczego skaningu laserowego (ALS). Analiza przydatności danych ALS bazuje na pomiarach referencyjnych 432 sosen w obszarze badawczym Milicz (RDLP Wrocław). Osiągnięte wyniki wskazują, iż: (1) wysokość pojedynczych drzew określona na podstawie modelowanej chmury punktów ALS w zależności od użytego algorytmu, prowadzi do niewielkiego zaniżenia wartości (średnia różnica -0.90 m CHM1 lub -0.12 m CHM2); (2) średnia wysokość analizowanych drzewostanów na powierzchniach kołowych określana na podstawie ALS, była wyższa (+0.85 m) od wartości zapisanych w bazie danych SILP (2005) co w efekcie prowadzić może do zaniżenia zasobności drzewostanów w całym obrębie; (3) automatycznie analizowana średnia wysokość drzewostanów sosnowych (95 centyl FE) była wyższa o +0.46 m w stosunku do danych z SILP; (4) automatycznie określona podstawa korony sosny wykazuje błąd zaledwie około 0.56 m; (5) analiza histogramów chmury punktów umożliwiła określenie długości korony z przeszacowaniem o +0.44 m w stosunku do danych referencyjnych; (6) modelowanie warstwy koron otwiera dyskusje na temat homogeniczności jednowiekowych i jednogatunkowych pododdziałów oraz istniejącego podziału przestrzennego. Nowa metoda inwentaryzacji leśnej bazująca na technologii pozyskiwania danych ALS i modelowaniu chmury punktów ma duże szanse na wdrożenie w lasach Polski o powinno przynieść ze sobą wzrost dokładności jak i obniżyć koszty prac urządzeniowych.

EN

The study was aimed at determining the utility of and to improve the understanding of the airborne laser scanning (ALS) technology in acquisition of selected parameters of canopy layers for individual trees and whole stands. This approach, based on ALS data (TopoSys fiber scanner; swing mode + optical line scanner), was compared with reference data drawn from a forest inventory (432 Scots pines). The study showed that: (1) the height of a single tree, as derived from ALS data, leads to underestimation (mean difference -0.90 m or +0.12 m depending on CHM generation algorithm); (2) the mean stand height was higher (+0.85 m) than the height recorded in the SILP database, which may result in underestimation of the timber volume in the entire Milicz forest district; (3) the stand mean height (understood as 95th percentile of the FE point cloud) was +0.46 m higher than the height recorded in the SILP inventory database; (4) it was possible to estimate the canopy base with 0.56 m overestimation; (5) the canopy length as measured during the forest inventory was +0.44 m lower compared to the ALS data (histogram analysis); (6) the homogeneity of a contemporaneous-pine stand is questionable. In a very near future, a new approach to the forest inventory, supported by ALS data, will be presented as a list of new parameters and guidelines.

9

Wykorzystanie technologii naziemnego skaningu laserowego w inwentaryzacji lasu

Chirrek M., Strzeliński P., Wencel A., Zasada M., Zawiła-Niedźwiecki T.

Roczniki Geomatyki

|

2007

|

T. 5, z. 5

19-24

EN

Terrestrial laser scanning (TLS) is the technology of the last few years. Within wide areas of its application, a big potential has been found for scientific and operational use of this technique to register and measure various characteristic features of forest environment. In Poland, the technology has been applied, among others, within the framework of the research project entitled .Forest inventory based on integration of various different geomatics techniques., commissioned by the General Directorate of State Forests to be implemented in the years 2006.2008. The goal of the project is to select methods of remote data collection suitable for forestry uses and to elaborate a forest inventory method based on selected geomatic techniques, including TLS technology. FARO LS 880 scanner was selected for the project. The main research area consists of sample plots located in the Milicz forest district. Consecutive tests were performed in Sławno, Chojna, Zielonka and Grodzisk forest districts. Tests of various variants of data collection and processing were performed. The obtained results concerning measurement accuracy and the scope of collected data were compared with traditional forest inventory methods.

10

Nowe kierunki pomiaru lasu z wykorzystaniem narzędzi teledetekcyjnych

Zawiła-Niedźwiecki T., Miścicki S., Zasada M., Wencel A.

Roczniki Geomatyki

|

2006

|

T. 4, z. 4

155-166

EN

The paper discusses methodological aspects of a research project financed by the State Forests dealing with forest inventory with the use of modern geomatic technologies (terrestrial and airborne laser scanning, hemispherical photographs, LAI, Mobile Mapping System, very high resolution satellite images).

11

Techniki teledetekcyjne w inwentaryzacji urządzeniowej lasu

Zajączkowski G., Wężyk P.

Roczniki Geomatyki

|

2004

|

T. 2, z. 4

41-50

EN

In the paper, preliminary results of research are presented. The project was executed by the Forest Management and Monitoring Unit of the Forestry Research Institute in cooperation with the GIS and Remote Sensing Laboratory at the Forest Ecology Department of the Forestry Faculty at the Kraków Agricultural Academy and was commissioned by the Directorate General of State Forests. Analysis of current state of the issue as discussed in world literature, not including results of the research of the authors. of this paper, was ordered by Forest Management Unit of the Forestry Research Institute. The aim of this research is assessment of currently available information sources about forests (alternative to traditional ground measurements) from the point of view of their possible use in forest inventory. Different remotely sensed data were assessed, including high-resolution Quickbird satellite images, panchromatic 1:13.000 aerial photos, AISA hyperspectral scanner images, DGPS measurements and others selected materials. Possible use of these sources of information on tree stands was tested in summer 2003 and 2004 in Staszów forest district (Regional Directorate of State Forests Radom) and in Niepo3omice forest district (Regional Directorate of State Forests Kraków). Usefulness of the following techniques was proved: 1) DGPS receivers and aerial and satellite images (QuickBird and aerial photos) in updating delineation of any internal boundaries in forest complexes; 2) registered digital images for acquiring information on crown closure of certain types of tree stands, species composition, number of trees and their spatial pattern; 3) laser scanning for acquisition of accurate data on tree height and vertical architecture of tree stands and landform features; 4) remote sensed data for evaluation of tree health status and for monitoring its changes in subsequent years of implementation of economic plans of forest regions; 5) combination of the a/m techniques for creation of a numerical model of land surface. It is forecasted that with the development of radar and laser techniques it will be soon possible to acquire precise information on the wealth of tree stands. In the conclusions: 1) attention was drawn to: a) limitations and flaws of these techniques such as cloudless weather and appropriate vegetation period for aerial and satellite photos, b) high price of QuickBird and IKONOS satellite images, expecting at the same time promotional prices, when IKONOS satellite receiving station is opened in Komorów near Warsaw; c) too small group of specialists in the area of processing and interpretation of aerial and satellite images in Poland; 2) the need for wider than up till now application of remote sensing techniques for acquisition of information for forest inventory was emphasized, first of all due its objectivity and usually higher accuracy as compared to traditional forest inventory practices.