Misganu Debella-Gilo

Sjefingeniør

(+47) 974 80 381
misganu.debella.gilo@nibio.no

Sted
Ås - Bygg R9

Besøksadresse
Raveien 9, 1430 Ås

Sammendrag

The size and location of agricultural fields that are in active use and the type of use during the growing season are among the vital information that is needed for the careful planning and forecasting of agricultural production at national and regional scales. In areas where such data are not readily available, an independent seasonal monitoring method is needed. Remote sensing is a widely used tool to map land use types, although there are some limitations that can partly be circumvented by using, among others, multiple observations, careful feature selection and appropriate analysis methods. Here, we used Sentinel-2 satellite image time series (SITS) over the land area of Norway to map three agricultural land use classes: cereal crops, fodder crops (grass) and unused areas. The Multilayer Perceptron (MLP) and two variants of the Convolutional Neural Network (CNN), are implemented on SITS data of four different temporal resolutions. These enabled us to compare twelve model-dataset combinations to identify the model-dataset combination that results in the most accurate predictions. The CNN is implemented in the spectral and temporal dimensions instead of the conventional spatial dimension. Rather than using existing deep learning architectures, an autotuning procedure is implemented so that the model hyperparameters are empirically optimized during the training. The results obtained on held-out test data show that up to 94% overall accuracy and 90% Cohen’s Kappa can be obtained when the 2D CNN is applied on the SITS data with a temporal resolution of 7 days. This is closely followed by the 1D CNN on the same dataset. However, the latter performs better than the former in predicting data outside the training set. It is further observed that cereal is predicted with the highest accuracy, followed by grass. Predicting the unused areas has been found to be difficult as there is no distinct surface condition that is common for all unused areas.

Sammendrag

Rapporten utforsker og diskuterer potensialet for økt bruk av Stordata (engelsk: big data) teknologi og metode innenfor instituttets arbeidsområder. I dag benyttes Stordata-tilnærminger til å løse forvaltningsstøtteoppgaver, samt til forskningsformål, særlig i sentrene for presisjonslandbruk og presisjonsjordbruk. Potensialet for økt bruk av Stordata innenfor instituttet er stort. For å realisere potensialet er det behov for god samordning mellom organisasjonsenhetene og utvikling av strategisk kompetanse på fagområdet.

Sammendrag

Rapporten dokumenterer utvalgte eksempler på bruk av stordata (engelsk: big data) teknologi og metode i NIBIO. Det første eksemplet er knyttet til oppdatering av arealressurskartet AR5, hvor det undersøkes om stordata-tilnærming kan benyttes for å identifisere lokaliteter der kartet må oppdateres. De neste eksemplene er hentet fra fagområdet plantehelse og tar for seg mulighetene for å bruke stordata-metode for å bedre prediksjonsmodeller og gjenkjenning av for skadegjørere.

Sammendrag

Planning sustainable use of land resources and environmental monitoring benefit from accurate and detailed forest information. The basis of accurate forest information is data on the spatial extent of forests. In Norway land resource maps have been carefully created by field visits and aerial image interpretation for over four decades with periodic updating. However, due to prioritization of agricultural and built-up areas, and high requirements with respect to the map accuracy, forest areas and outfields have not been frequently updated. Consequently, in some part of the country, the map has not been updated since its first creation in the 1960s. The Sentinel-2 satellite acquires images with high spatial and temporal resolution which provides opportunities for creating cloud-free mosaic images over areas that are often covered with clouds. Here, we combine object-based image analysis with machine learning methods in an automated framework to map forest area in Sentinel-2 mosaic images. The images are segmented using the eCogntionTM software. Training data are collected automatically from the existing land resource map and filtered using height and greenness information so that the training samples certainly represent their respective classes. Two machine learning algorithms, namely Random Forest (RF) and the Multilayer Perceptron Neural Network (MLP), are then trained and validated before mapping forest area. The effects of including and excluding some features on the classification accuracy is investigated. The results show that the method produces forest cover map at very high accuracy (up to 97%). The MLP performs better than the RF algorithm both in classification accuracy and in robustness against inclusion and exclusion of features.

Sammendrag

The development of a new bioeconomy implies an increased need for renewable biological resources. This means that more of the existing biomass will be har vested and a larger land area is likely to be utilized. While it is widely acknowl edged that this increased harvest and production must be sustainable, it must also be acknowledged that there are some potential challenges. For example, there may be different aims targeting the same area. To meet this challenge in an informed approach, we argue that geographical data and spatial analyses are key. We exemplify this through a study in which we utilize the spatial distribution of produced biomass from agriculture and forestry, together with the location of threatened species from the Norwegian Red List of Species. In our analyses we demonstrate that there is a spatial overlap between the most productive land for forestry and agriculture. At the same time, a high occurrence of threatened spe cies is also found in these areas. We conclude that analysing the geography of conflicting aims is important. It documents the importance of spatial data, and findings from this type of analyses need to be included in bioeconomy decision making.

Sammendrag

Regjeringens bioøkonomistrategi, Kjente ressurser – uante muligheter, fra 2016, fremhever at den fremtidige bioøkonomien må være bærekraftig. Når en større del av våre behov skal dekkes med utgangspunkt i fornybare biologiske res¬surser, vil vi måtte høste mer av disse ressursene – selv om det også er et mål å utnytte ressursene bedre. Disse ressursene skal imidlertid høstes på det samme landarealet som utgjør leveområder for sjeldne arter og verdifulle naturtyper. Og disse har vi også forpliktelser overfor, gjennom FNs bærekraftsmål og konvensjonen om biologisk mangfold. Her ligger det en mulig konflikt mel¬lom ulike interesser. Hvordan kan vi øke og effektivisere uttaket av fornybare biologiske ressurser på en måte som ikke kommer i konflikt med målet om å sikre naturmangfoldet og målet om en miljømessig bærekraftig bioøkonomi? I denne sammenhengen er det nødvendig å være oppmerksom på at de land¬baserte fornybare biologiske ressursene til en viss grad konkurrerer om det samme arealet. Og det kan være slik at å produsere mer av én ressurs reduserer muligheten for å produsere en annen. Det finnes heller ikke noen omforent fasit for hva som er en miljømessig bærekraftig løsning. Dette er komplekse sammenhenger, og svaret kan vi ikke få gjennom noe enkelt regnestykke. Men ved hjelp av geografiske analyser kan vi belyse noen av de mulige utfordrin¬gene. I dette kapitlet legger vi frem analyser vi har gjort av forholdet mellom areal og potensielt økt uttak av skog. Vi har også analysert hvordan et slikt uttak kan komme til å påvirke rødlistede arter. Avslutningsvis ser vi nærmere på arealkonflikter som kan oppstå dersom skogarealet økes. Målet med analy¬sene er å bidra til en mest mulig kunnskapsbasert forvaltning av de landbaserte fornybare biologiske ressursene, samtidig som vi kommer nærmere målet om å sikre en bioøkonomi som også er miljømessig bærekraftig.

Til dokument

Sammendrag

A transition to a bioeconomy implies an increased focus on efficient and sustainable use of biological resources. A common, but often neglected feature of these resources is their location dependence. To optimize their use, for example in bioeconomic industrial clusters, this spatial aspect should be integrated in analyses. Optimal design and localization of a bioeconomic cluster with respect to the various biological and non-biological resources required for the cluster, the composition of industrial facilities in the cluster, as well as the demands of the outputs of the cluster, is crucial for profitability and sustainability. We suggest that optimal design and location of bioeconomic clusters can benefit from the use of a Multicriteria Decision Analysis (MCDA) in combination with Geographic Information Systems (GIS) and Operations Research modeling. The integration of MCDA and GIS determines a set of candidate locations based on various criteria, including resource availability, accessibility, and usability. A quantitative analysis of the flow of resources between and within the different industries is then conducted based on economic Input-Output analysis. Then, the cluster locations with the highest potential profit, and their composition of industrial facilities, are identified in an optimization model. A case study on forest-based bioeconomic clusters in the Østfold county of Norway is presented to exemplify this methodology, the expectation being that further implementation of the method at the national level could help decision makers in the planning of a smoother transition from a fossil-based economy to a bioeconomy.

Sammendrag

Nation-wide Sentinel-2 mosaics were used with National Forest Inventory (NFI) plot data for modelling and subsequent mapping of spruce-, pine-, and deciduous-dominated forest in Norway at a 16 m × 16 m resolution. The accuracies of the best model ranged between 74% for spruce and 87% for deciduous forest. An overall accuracy of 90% was found on stand level using independent data from more than 42 000 stands. Errors mostly resulting from a forest mask reduced the model accuracies by ∼10%. The produced map was subsequently used to generate model-assisted (MA) and poststratified (PS) estimates of species-specific forest area. At the national level, efficiencies of the estimates increased by 20% to 50% for MA and up to 90% for PS. Greater minimum numbers of observations constrained the use of PS. For MA estimates of municipalities, efficiencies improved by up to a factor of 8 but were sometimes also less than 1. PS estimates were always equally as or more precise than direct and MA estimates but were applicable in fewer municipalities. The tree species prediction map is part of the Norwegian forest resource map and is used, among others, to improve maps of other variables of interest such as timber volume and biomass.

Sammendrag

This paper describes the development and utility of the Norwegian forest resources map (SR16). SR16 is developed using photogrammetric point cloud data with ground plots from the Norwegian National Forest Inventory (NFI). First, an existing forest mask was updated with object-based image analysis methods. Evaluation against the NFI forest definitions showed Cohen's kappa of 0.80 and accuracy of 0.91 in the lowlands and a kappa of 0.73 and an accuracy of 0.96 in the mountains. Within the updated forest mask, a 16×16 m raster map was developed with Lorey's height, volume, biomass, and tree species as attributes (SR16-raster). All attributes were predicted with generalized linear models that explained about 70% of the observed variation and had relative RMSEs of about 50%. SR16-raster was segmented into stand-like polygons that are relatively homogenous in respect to tree species, volume, site index, and Lorey's height (SR16-vector). When SR16 was utilized in a combination with the NFI plots and a model-assisted estimator, the precision was on average 2–3 times higher than estimates based on field data only. In conclusion, SR16 is useful for improved estimates from the Norwegian NFI at various scales. The mapped products may be useful as additional information in Forest Management Inventories.

Til dokument

Sammendrag

Aim: Identify where bioeconomic development would best be located to maximise both local resources and the reusable waste from potentially collaborating sectors. We seek to answer the questions like Where are the best locations for bioeconomic clusters and how should this be assessed? What are the tradeoffs, how can they be mapped and described, and are there any general major obstacles? What are the conditions that would aid in developing a smart bioeconomy and what are the spatial implications of different developments?

Til dokument

Sammendrag

In 2012, the Norwegian Environmental Agency funded an extension to the Global Pollination Project, coordinated by the FAO (Food and Agriculture Organization of the United Nations) to expand the number of connected countries from 7 fully participating to in total 13 countries. This international effort seeks to build capacity for pollination studies and add to the knowledge base for the Intergovernmental Platform for Biodiversity and Ecosystem Services (IPBES). IPBES is currently conducting its first fast track case study on pollination. Specifically, the Global Pollination Project implements the “Protocol to detect and assess pollination deficits in crops: a handbook for its use” (Vaissière et al. 2011), developed through the FAO. The proto-col outlines a unified method to investigate pollination and measure pollination deficits in vari-ous agricultural systems around the world. NINA (the Norwegian Institute for Nature Research) was tasked with setting up a Norwegian collaboration to implement the protocol in Norway, to analyse its applicability to Nordic conditions and evaluate its strength in relation to alternative research strategies. The present report is the result of this effort. This report does not communicate the final results of the analyses, as they will be conducted in the two larger “host-projects” that made the implementation of the protocol possible. Instead, it outlines the rationale of the protocol, and evaluates its potential for providing management rel-evant information on pollination deficits, with particular emphasis on Norway. We discuss the state and trends of pollination dependent crops in Norway, as a background for the need for pollination in Norwegian Agriculture. The protocol is general enough to be applied to a wide variety of settings, and we did not expe-rience any fundamental problems of implementing it in a Nordic setting. We did however notice challenges to an effective implementation, which might be especially pronounced in a Norwe-gian or Scandinavian setting. First, it can be difficult to find a wide enough range of factors that influence pollinators to efficiently analyse the importance of pollination without resorting to ma-nipulative treatments. For example, the amount of flower resources and fragmentation of habi-tat are factors known to influence pollinators. But many crops are spatially aggregated to rela-tively narrow valleys and therefore experience similar surroundings. Secondly, it can be chal-lenging to find enough replicate farms since Norway is a relatively small agricultural nation. Thirdly, pollinators in Norway (as in many other parts of the world) are intractably linked to ag-ricultural and animal husbandry practices that provide a diversity of flowering resources neces-sary for pollinating insects, yet these practices and resulting resources in the surrounding land-scape is not sufficiently captured by the survey protocol. The protocol is designed to estimate differences in yield given differences in pollination, and various methods are available to approach optimal pollination, that acts as benchmark. Esti-mating the effect of pollination on yield is the foundation to understanding the status of pollina-tion deficits for any crop. The protocol appears to be a successful effort to create a unified standard of measuring pollination and pollination deficits by this definition. It thus marks a great improvement for pollination research in agriculture internationally. Pollination, Ecosystem services, Bees, Bumble bees, Pollination deficit Protocol, FAO, IPBES, Policy, apple, red clover, Norway, pollinering, økosystemtjenester, bier, humler, protokoll for polline-ringsunderskudd, FAO, IPBES, eple, rødkløver

Sammendrag

The Norwegian landscape is changing as a result of forest regeneration within the cultural landscape, and forest expansion has impacts on accessibility, visibility, and landscape aesthetics, thereby affecting the country's tourism industry. This study aimed at identifying the potential areas of forest regeneration and anticipated subsequent landscape effects on different categories of tourist locations in southern Norway. Deforested areas with a potential for forest regeneration were identified from several map sources by GIS-analyses, and 180 tourist locations were randomly selected from the Norwegian national tourism database (Reiselivsbasen), and then buffered by 2 km radius for land cover classes. The findings revealed that approximately 15% of southern Norway has the climatic potential for future forest regeneration, in addition to 5% of cultivated land. Future forest regeneration will affect the landscapes surrounding the tourist locations of rural south Norway, and while the potential is nationwide, it is not uniformly distributed. Two important tourist landscape regions seem especially exposed to forest regeneration: the coastal heath region and the mountain landscapes. Large parts of these areas do not have sufficient numbers of domestic grazing animals necessary to maintain the present character of the landscape.

Sammendrag

Reiselivet i Norge har de siste tiåra vært oppmerksomme på landskapsendringene som skjer i Norge. I følge reiselivsnæringa truer gjengroing av kulturlandskapet viktige segmenter innen det norske reiselivet. Samtidig legges det årlig ned et sted mellom 1500 - 2000 gardsbruk i Norge.

Sammendrag

En produksjonslinje for publisering av jordsmonndata og annen relevant jordbruksstatistikk per vassdragsenhet (REGINE-enhet) er etablert. Her presenteres statistikken for Drammensvassdraget. Informasjonskilden for denne rapporten er først og fremst jordsmonndatabasen, men det er også lagt inn noe informasjon fra produksjonsregisteret (temaene planteproduksjon og dyretetthet). Jordsmonndatabasen dekker kun dyrket mark og bare den sydlige halvdelen av vassdraget. Jordsmonnstatistikken er utarbeidet for vassdragsenheter på overordnet nivå som har mer enn 80 % dekning av jordsmonndata. Kartlagt areal er ikke representative for vassdraget som helhet. Det er også utarbeidet en mer detaljert statistikk for følgende enheter: Vigga-Randsfjorden i Oppland, Simoa i Buskerud og deler av Vestfossenelvas nedslagsfelt i Vestfold. Her er det også laget statistikk som viser hvordan egenskapene varierer med avstanden fra vannstrengen.

Sammendrag

Reiselivets landskap er vurdert m.h.p. gjengroing av kulturlandskap. Særlig utsatt for gjengroing av utmark er reiselivskategoriene gardsturisme og hoteller. Seterregionene og kystregionene vil i framtiden være spesielt utsatt for gjengroing av utmarkas kulturlandskap.