GIS-based Spatial-Temporal analysis of the biomass production in the Republic of Srpska teritory

My first co-author (Bajić, D., Adžić, D., Dekić, R.) Remote Sensing Analysis Originally Published in Herald

Spatial-temporal analysis of the biomass production in the Republic of Srpska territory by using products of remote sensing and geographic information technologies. The basic input parameters on which the analysis is based are Gross Primary Production (GPP) and Net Primary Production (NPP). Based on GPP and NPP as measures of carbon production and accumulation in ecosystems, the biomass production was measured indirectly. MODIS (Moderate Resolution Imaging Spectroradiometer) MOD17 product-based satellite footage was used as it generates eight-day and yearly intervals of GPP and NPP parameters as GIS raster layers. The analysis was performed for the 2005-2014 ten-year period by using Google Earth Engine web GIS application and QGIS desktop GIS application. Results of the biomass production analysis in the Republic of Srpska territory during the target period indicate specific deviations and oscillations within spatial and temporal frameworks, depending on the ecosystem type and weather conditions during some of the observed years.


The measure of an ecosystem efficiency is the size of its organic production. It is expressed with the amount of organic matter synthesized in a unit of time over the unit of surface/volume (Pešić, 2011). The land primary production provides energy for the maintenance of an ecosystem’s structure and function and delivers resources such as food, wood, fuel and fibres crucial for human society. The biomass production may be indirectly determined in several different manners such as follows: measuring the quantity of released oxigen, measuring the quantity of

used carbon dioxide, or measuring the carbon flow. The paper performs the analysis of biomass production by measuring the carbon flow and accumulation within ecosystems. In order to detect the primary mechanism for the biomass production measurement, we provide an outline and definition of primary terms and the bibliography in which these terms are defined. The primary production is the rate of solar energy converted into biomass by the processes of photosynthesis or homeosynthesis. The total energy converted by photosynthesis is called gross primary production (GPP). The net primary production is a result of difference between the gross primary production and the energy spent during the breathing process (Campbell, 1990). The net primary production is regarded as a net amount of carbon (C) stored in plants and accumulated as biomass.

NPP = GPP-Respiration

Measurement of the net primary production is crucial once the energy flow within an ecosystem can be measured, after which it is possible to provide a description of the ecosystem condition and predict a response to changes and deviation based on the measurement results (Whitten, et al., 1996). The reasearch on net primary production is pertinent as it is a key process in carbon circulation and it directly measures the production within an ecosystem (Zhao and Running, 2008).As a tool for measurement of production and amount of accumulated carbon, the net primary production may be used for the evaluation of impact of natural and artificial (human) divergence on the productivity within an ecosystem. Globally, the land net primary production is one of the mostly modelled ecological parameters with models that largely differ in their approach and complexity, often providing similar comparative evaluation (Field et al., 1995). It is crucial to uderstand the relation between the net primary production and carbon circulation in order to predict and assuage negative environmental changes. The annual NPP or the accumulated amount of carbon assimilated by biomass may be used for the estimation of plant growth, i.e. biomass (Jihye, et al., 2011).

Apart from the fact that it is not possible to perform direct measurements, there are many models based on physiological principles and global Earth dynamics (Cramer et al. 1999).The estimation of carbon-dioxide (CO2) exchange between biosphere and atmosphere as well as vegetation production on regional, continental and global levels may be achieved by combining remote sensing and the information on the carbon circulation process. The net primary production considers how much CO2 is absorbed by vegetation during photosynthesis and how much CO2 is emitted during breathing, which is a process of plants consuming sugar for energy production. The net amount of CO2 absorbed by vegetation may be converted into dry matter – biomass (FAO, 2015). NPP refers to the net amount of carbon absorbed by vegetation as a result of assimilation (inward) and respiration (outward). The amount of C is expressed as the mass per surface unit (C kg/m2). The CO2 flow is a source of carbon within plants and the general photosynthesis equation determines if the carbon converts into carbon-hydrates (CH2O), which makes it a ket element of the vegetation dry matter. Depending on the molecular mass, C mass may convert into the production of dry matter (CH2O) or the production of dry biomass (FAO, 2015).The analysis of biomass production capitalizes on the satellite data as they provide mechanisms for the assessment, monitoring and evaluation of spatial and temporal production within land ecosystems (Crabtree et al. 2009). For the evaluation of NPP, our study uses the following MODIS satellite products: MOD17A3H (MODIS/Terra Net Primary Production Yearly L4 Global 500 m SIN Grid V006) and MYD17A2H (MODIS/Aqua Gross Primary Productivity 8-Day L4 Global 500 m SIN Grid V006). Based on the defined satellite products, we generated annual NPP indicators as GIS raster maps, separately for each year during the target 2005-2014 ten-year period. These generated maps were the basis, i.e. input information, for the model of spatial-temporal analysis of biomass production. Apart from the maps, we generated annual tables with information on temporal NPP distribution on the 8-day level within one year. The collection, processing and adjustment of the satellite products were performed via Google Earth Engine (GEE), which enables access and analysis of a large number of sets of spatial data. GEE is a recent technology which represents a web-based platform intended to monitor and measure changes in Earth’s surface. The platform offers free access to an extensive data catalogue resulting from years of Earth observation. GEE calculationsare performed via Google infrastructure, in which analyses are automatically parallelized so that specific calculations may parallelly cover multiple computer processors. The GEE data catalogue comprises 5-bit data on Earth observations and the application itself, so as a result, we elliminate data download and management on a local infrastructure. It is crucial for extensive and demanding geospatial analyses. GEE stores source data in its original projections along with all original information and metadata. Data from other sources may also be used and reprojected in GEE if necessary.


The data used in this paper are based on NASA remote sensing products downloaded from TERRA MODIS and AQUA MODIS sensors. NASA’s MOD17 project has been providing continuous information on GPP and NPP for the whole Earth surface, starting with 2004. The objective of MOD17 is to continuously evaluate gross/net primary productions of the entire Earth vegetation. The outputs of MOD17are usefull for the management of land and natural resources, the global analysis of carbon circulation, monitoring of environment changes, and assessment of ecosystem status.We used the following MODIS products for calculating and analyzing the net primary production: annual cumulative NPP MOD17A3H, and MODIS-based MYD17A2H product which represents a cumulative indicator of gross primary production (GPP) for each eight days. Both these products have the spatial resolution of 500 m.

The NTSG (Numerical Terradynamic Simulation Group) research laboratory at the University of Missoula, Montana (USA) provides continuous assessment of GPP and NPP based on MODIS data. These data are a result of the MOD17 algorithm based on the original Monteith logic (Monteith, 1972) on the efficiency of radiation usage.

In GEE (Googl Earth Engine), two MODIS products were extracted from the data catalogue: MOD17A3H (annual NPP) and MYD17A2H (GPP cumulative for 8-day period) for a predetermined observation period (2005-2014). In order to analyse the annual shift of NPP values, each 8 days more precisely, we calculated NPP for each 8 days. The cumulative NPP for 8 days takes into consideration the maintenance, growth, and breathing of plants during a year and it it covers the data set in line with GPP values for each 8 days and the sum of all GPP values during a year (annual GPP). Calculation of NPP 8 was performed in line with the following formula:NPP 8=( GPP8/ GPPyear) x NPPyear NPP8 values during a year were extracted from GEE in Excel format, in which they are presented in forms of diagrams. NPP describes the net amount of carbon (C) absorbed by vegetation as a result of assimilation (inward) and respiration (outward). The amount of C is described as mass per surface unit (kg/m2). The carbon-dioxide (CO2) flow is the plant carbon source, and the general photosythesis equation determines if the carbon converts into carbon-hydrates (CH2O), thus being the key element of the vegetation dry matter. Considering its molecular weight, C weigth may convert into dry matter production (CH2O) or

biomass production (FAO, 2015). Conversion of carbon into dry matter or dry biomass was performed in line with the following formula (the amount of accumulated carbon

multiplies with the molecular mass of carbon-hydrates) (FAO, 2015):

Biomass=NPP x 30/12

Values of biomass production are expressed in kg/m2of dry biomass and generated as raster GIS layers in GEE application. Due to a thorough processing, the layers were exported in .tif format. Once they were downloaded, processed and exported by using GEE application, the data were entered in the QGIS desktop GIS application and the analysis of spatial-temporal features of biomass production in the Republic of Srpska was performed.


Diagram 1 displays the average annual dry biomass production in the Republic of Srpska expressed in kg/m2 during the observed ten-year period. The diagram shows that the biomass production during the observed period varied from 1,48 to 1,98 kg/m2. Since the biomass production directly depends on climate conditions, we may assume that in the dry years (years with poor precipitation), the biomass production was lower. Hence, the years of 2011 and 2012 are specific.

Diagram 1 – Average annual dry biomass production in the Republic of Srpska (2005-2014)

Map 1 displays the spatial distribution of average annual dry biomass production in the Republic of Srpska during our target ten-year period. The spatial distribution of biomass production depends on several factors, and the most pertinent ones are land use and climate conditions. The map indicates that biomass production is more pronounced in forest areas, especially in zones of deciduous woods. Due to their physiology, conifer woods produce somewhat less biomass. In zones of intensive agrarian production, there is a low biomass production. This observation may be accounted for by the fact that, in natural ecosystems, the production takes place during the whole vegetation period, whereas in artificial ecosystems (arable land) the production takes place only during a part of vegetation period (from sieving until harvest).

Map 1 – Spatial distribution of average annual dry biomass production in the Republic of Srpska (2005-2014)

Diagram 2 displays a temporal distribution of dry biomass production during specific observed years. We may clearly infer that the intensive biomass production in the Republic of Srpska takes place during vegetation period (March-October),which is directly conditioned by weather conditions (temperatures). During specific years, we observe slight deviations in production in summer season, which is directly affected by precipitation regime. Furthermore, we may clearly conclude about the start of intensive production. Depending on the temperature in spring, the intensive production starts in early March/late April.

Diagram 2 – Temporal distribution of average dry biomass production per years during the observed period (kg/m2)

Map 2 displays deviations of annual dry biomass production from the average annual production during the observed ten-year period, for each year individually. The map shows positive (production larger than average) and negative (production smaller than average) deviations. Hence, we may clearly see the temporal and spatial distribution of the defined deviations. Years

2010 and 2014 are the years of the above average biomass production in the Republic of Srpska. On the other hand, year 2012 is the year of the below average biomass production in the whole target territory. In addition, 2011 was the year of low production in north Republic of Srpska and 2008 was the year of low production in south of the territory.

Map 2 – Positive and negative deviations of the annual biomass production in contrast with the average ten-year production


Spatial-temporal research on the biomass production are highly pertinent and applicable in different aspects of human actions such as the ecosystem condition monitoring, climate change monitoring, tracking the impact of climate changes on the ecosystem modification, decision making on the adjustment to climate changes, ecosystem management, ecosystem usage planning, etc. Modern information technologies enable an efficient access to performing spatial-

temporal analyses on biomass production, especially geographical information technologies and remote sensing that are crucial for data collection and analysis. The paper outlined and described the procedure of downloading and processing of remote sensing products, which were used to identify and analyze biomass production in the Republic of Srpska and finally, displayed the reasearch results. The observed ten-year period showed specific spatial and temporal deviations in distribution of dry biomass production in the Republic of Srpska territory. These deviations may be connected with climate conditions (temperatures and precipitation), land use, and health ballance of specific ecyststems. The methodology applied in the paper and the results displayed may be of both theoretical and practical use within the fields of ecology, environment protection, climatology, agrarian production, forestry, spatial planning, etc.


Zhao,M.,Running,S.W.,2008.Remote sensing of terrestrial primary production and carbon cycle.

In:Liang,S.(Ed.),Advances in Land RemoteSensing. Springer Science Business Media, New York, pp. 423–444, ISBN:978-1-4020-6449-4

Campbell, N. A. (1990). Biology. Redwood City, CA, TheBenjamin/Cummings Publishing Company, Inc.

Crabtree, R., Potter, C., Mullen, R., Sheldon, J., Huang, S., Harmsen, J., Rodman, A., Jean, C. (2009). A modeling and spatio-temporal analysis framework for monitoring environmental change using NPP as an ecosystem indicator. Remote Sensing of Environment 113 (2009) 1486–1496. Elsevier Science Inc., 2009.

Cramer, W., and C. B. Field. 1999. Comparing global models of terrestrial net primary productivity (NPP): introduction. Global Change Biology 5(Supplement 1):iii–iv.

FAO project TCP /AFG/3402, 2015. Water accounting through Remote Sensing (WA+) in Helmand River Basin. Analysis on water availability and uses in Afghanistan river basins.

Field,C. B., Randerson, J. T.,and Malmstrom, C. M. (1995). Global net primary production: Combining ecology and remote sensing. Remote Sensing of Environment. 51:74-88 (1995) Elsevier Science Inc., 1995.

Hansen, M., Potapov, P., Moore, R., Hancher, M., Turubanova, S., Tyukavina, A., Thau,D., Stehman, S., Goetz, S., Loveland, T., et al., 2013. High-resolution global maps of 21st-century forest cover change. Science 342 (6160), 850–853.

Jihye, L., Sinkyu, K., Keunchang, J., Jonghan, K., Sukyoung, H. (2011). Monitoring of Gross Primary Productivity (GPP) and Crop Yield via MODIS in Soyang River Basin, South Korea. 2011 TERRECO Science Conference October 2 – 7, 2011; Karlsruhe Institute of Technology, Garmisch-Partenkirchen, Germany.

Whitten T, Whitten A, Soeriaatmadja R, Afiff S (1996) Theecology of Java and Bali. Oxford University Press,Oxford

Zhao, M., & Running, S. W. (2008). Remote sensing of terrestrial primary production and carbon cycle. In М Advances in Land Remote Sensing (pp. 423-444). Springer Netherlands.Zhao, T., Brown, D.

Zhao, M., & Running, S. W. (2015). User’s Guide Daily GPP and Annual NPP (MOD17A2/A3) Products NASA Earth Observing System MODIS Land Algorithm. Version 3.0 For Collection 6 October 7, 2015. NASA DAAC.

Pešić, B., S., (2011) Osnovi ekologije. Univerzitet u Kragujevcu, Prirodno-matematički fakultet