Predictive Modelling of Aquatic Ecosystems at Different Scales using Mass Balances and GIS

Full text

(1)Comprehensive Summaries of Uppsala Dissertations from the Faculty of Science and Technology 957. Predictive Modelling of Aquatic Ecosystems at Different Scales using Mass Balances and GIS BY. ANDREAS GYLLENHAMMAR. ACTA UNIVERSITATIS UPSALIENSIS UPPSALA 2004.

(2)

(3)

(4)

(5)

(6)

(7)

(8)

(9)

(10) ! " #"$ % & % % '& &( )&

(11)

(12) *

(13)

(14) +*&(

(15) & ( !( '

(16) % , - %%

(17) +

(18) .

(19)

(20) /+(

(21)

(22) (

(23)

(24)

(25)

(26)

(27)

(28) 0$1( $ ( ( /+.2 0"3$$!3$03 )& &

(29) % , ( & /

(30) %

(31) + 4/+5 %

(32)

(33) % & &( )&

(34)

(35) %

(36)

(37)

(38) % 6

(39)

(40)

(41)

(42) %%

(43)

(44) 4

(45)

(46)

(47)

(48)

(49) 5( )&

(50)

(51) %

(52)

(53)

(54) %

(55) % & &(

(56) *

(57) &

(58)

(59) % & &

(60)

(61)

(62) %

(63) %& %

(64) 4(( & 5( )& * *& %

(65)

(66)

(67) & , %

(68) ,

(69) (( 7

(70)

(71)

(72) 4

(73) 8 *

(74) &

(75) &

(76) 5(

(77) *

(78) %

(79) 4+'5

(80)

(81)

(82)

(83) 4(( & 5 * & %%

(84) & &

(85)

(86) & + ( )&

(87) & &

(88)

(89)

(90) & +. 4 &

(91)

(92)

(93)

(94)

(95) 5 % %

(96)

(97)

(98) % &

(99) ( % * % 8 *

(100) %

(101) % . +

(102)

(103) 4(( &

(104)

(105)

(106) 5( )&

(107)

(108) % & &

(109) .

(110) % 8

(111)

(112)

(113) % & &

(114) %

(115)

(116) % 8 %

(117)

(118) %

(119) ( )&

(120) % &

(121) * %

(122) %& , % *

(123) ( 9 &

(124)

(125) 4((

(126) 5 &

(127) %% %

(128)

(129) ( )&%

(130) *

(131) *

(132)

(133) %

(134) 4:+./5( )& & * % & 9

(135)

(136) & & +.

(137) & ;8

(138)

(139) . *

(140) (

(141) % 3

(142)

(143)

(144) & & * &

(145)

(146) * * & & :+./ & (

(147)

(148)

(149)

(150) %& %

(151) *

(152)

(153)

(154) & &

(155)

(156) /+

(157)

(158)

(159)

(160)

(161) , % * %&

(162) , . + & + ;8

(163) .

(164) ! " #

(165) " $ % &'"

(166)

(167) " #()*+,'

(168) " <

(169)

(170) & ! /++2 "" !3=> /+.2 0"3$$!3$03

(171) #

(172)

(173) ### 3!"!= 4& #66

(174) (8(6 ?

(175) @

(176) #

(177)

(178) ### 3!"!=5.

(179) Jump in! Get your feet wet! Make a splash! Together we can make a difference. No matter who we are, where we are, and what we do, we are all dependent on water. We need it every day, in so many ways. We need it to stay healthy, we need it for growing food, for transportation, irrigation and industry. We need it for animals and plants, for changing colours and seasons. from the UNESCO official website of the International Year of Freshwater 2003. To my family and In memory of Per Lönn.

(180)

(181) List of papers included in this thesis. This thesis is based on the following papers, which in the comprehensive summary will be referred to by their Roman numerals: I. Gyllenhammar, A., Håkanson, L. and Lehtinen, K-J. 2004. Modelling nutrient dynamics in a mesocosm fish farming experiment. Submitted.. II. Håkanson, L., Gyllenhammar, A. and Brolin, A. 2003. A dynamic compartment model to predict sedimentation and suspended particulate matter in coastal areas. Ecological Modelling. In press.. III. Gyllenhammar, A. and Gumbricht, T. 2003. WASUBI: A GIS tool for subbasin identification in topographically complex waterscapes. Environmental Modelling and Software. Accepted.. IV. Håkanson, L. and Gyllenhammar, A. 2003. A new way of setting fish quotas based on holistic ecosystem modelling and accounting for key environmental factors and foodweb interactions regulating primary and secondary production. Submitted.. V. Gyllenhammar, A. and Håkanson, L. 2004. Environmental consequence analyses of fish farm emissions related to different scales and exemplified by data from the Baltic – a review. Submitted..

(182) The contributions to the papers by the author is as follows: I) II) III) IV) V). Responsible for the main part of the model development, analyses, evaluation and writing. Responsible for parts of data compilation and modelling. Contributed with ideas and discussion. Responsible for the main part of the model development, programming, analyses, evaluation and writing. Responsible for parts of data compilation and modelling. Contributed with ideas and discussions. Responsible for parts of the theory formulation and writing. Contributed with ideas and discussion..

(183) Contents. 1. Introduction and aim..............................................................................9. 2. Methods ...............................................................................................11 2.1 Mass balance modelling .............................................................11 2.1.1 Theory and background .........................................................11 2.1.2 Fish farming and the mesocosm model .................................12 2.1.3 The concept of functional groups ..........................................13 2.1.4 From LakeWeb to BaltWeb ...................................................14 2.1.5 Fish farming as a point source of nutrients............................14 2.2 Image segmentation....................................................................15. 3. Regional scale study areas ...................................................................17 3.1 The Archipelago Sea, Finland ....................................................17 3.2 The Okavango Delta, Botswana .................................................17. 4. Highlighted results and discussions.....................................................21 4.1 The mesocosm nutrient model (Paper I) ....................................21 4.2 The SPM model for coastal areas (Paper II)...............................23 4.3 The delineation of topographically complex waterscapes (Paper III) ...................................................................................24 4.4 The BaltWeb simulations (Paper IV) .........................................26 4.5 Environmental consequence analyses of fish farm emissions (Paper V) ....................................................................................27. 5. Main findings and conclusions ............................................................28. 6. Acknowledgements..............................................................................30. 7. Summary in Swedish ...........................................................................32. 8. References ...........................................................................................34. Appendix A. WASUBI source code.

(184)

(185) 1 Introduction and aim. In aquatic ecosystem studies, it is easy to be overwhelmed when one realizes that everything is related to everything else, and often in very complex webs or systems. However, as an aquatic modeller, there is a possibility to try to structure the knowledge and look for the relevance and importance of different biotic and abiotic processes. Society poses many questions about the environmental impact at different scales and scientists have to search for answers both for historical explanatory reasons and, maybe even more important, to predict possible future scenarios. This is a common denominator, regardless whether the concern is contamination, climate change, eutrophication, acidification or any other known or unknown environmental threat. The present thesis aims to study and improve the framework of aquatic ecosystem modelling at different temporal and spatial scales. In this thesis, sections 2-4 aim to give a summary of the work presented in Papers I-V and to give brief overviews, backgrounds and theories of previous work in the field. The five papers on which this thesis is based are connected by the aim to model aquatic ecosystems at different scales. They include studies of system dynamics and quantifications of nutrient flows at different scales (Papers I, II and IV), or summarise the results into a scale perspective (Paper V). For the intermittent scale (between site and international), an attempt is made (Paper III) to develop a basis for modelling improvement by the construction of an objective delineation method for topographically complex waterscapes. Paper I estimated the nutrient dynamics in a mesocosm fish farming experiment. This was done by developing a mass balance model for both nitrogen and phosphorus. After critical model testing with uncertainty and sensitivity analyses, a special focus was set on the rainbow trout (Oncorhynchus mykiss) growth modelling. One of the main goals of the study was to calculate which feeding conditions are required for a sustainable aquaculture scenario. This can be done by the inclusion of regionally caught wild fish, in this case Baltic herring or sprat from the Finnish Archipelago Sea. Consequently, the threshold value for achieving a zero nutrient loading at a larger scale can be calculated. Paper I also aims to illustrate the important link between the site scale and the regional scale.. 9.

(186) Paper II presents a new dynamic mass balance model for suspended particulate matter (SPM) and sedimentation in coastal areas (i.e. the local scale). The aim was to construct a model that handles all important fluxes of SPM to, from and within coastal areas. The coastal areas were defined by the topographical bottleneck method using GIS. In order to make the approach practically useful, the aim was to construct a model that may be run using variables readily accessible from coastal monitoring programs and/or topographic maps. In Paper III, a new method for delineation of topographically complex waterscapes is described. The aim was to develop an objective and automatic method that requires a minimum of input data and very little user interaction. The program, which was constructed as a set of scripts in the ArcView programming language Avenue, used a digital elevation model (DEM) to find local troughs in order to create a user defined number of subbasins. Two study areas, the Finnish Archipelago Sea and the Botswanan Okavango Delta, were used for model development and testing. Two of the main goals were to examine model sensitivity as for the quality and the resolution of the given data sets and to compare openness and homogeneity of the subbasins created by the model with randomly created subbasins. The aim of Paper IV was to examine a new way of setting fish quotas from holistic aquatic foodweb modelling at the international scale (for the Baltic Sea). To complement the methods commonly used today, which are based on fish catch statistics, this approach includes foodweb interactions. Furthermore, it accounts for changes in environmental conditions, e.g. salinity and temperature. The model is a transformation and recalibration for the Baltic Sea of an existing lake model, LakeWeb, and the results presented in Paper IV should be regarded as preliminary. The aim was also to present results on how environmental conditions influence the aquatic foodweb and estimate the sustainability of present cod fishing in the Baltic Sea. Paper V connects the other papers in this thesis and puts them into a wider context. The idea was to review how point source emissions of nutrients influence aquatic ecosystems at various scales. Fish farming was used as a common example of a point source. Four different scales in the Baltic Sea were used: (1) a fish farm site, (2) a local coastal area, (3) a regional area and (4) an international scale. Paper V also aimed at reviewing available literature concerning these matters.. 10.

(187) 2 Methods. 2.1 Mass balance modelling 2.1.1 Theory and background No matter what reactions a substance may undergo and regardless of its phase distribution, the following statement is always true: ªInput rate º ªOutput rate º ªNet transformation º ªAccumulation or º » « depletion of X » (1) « of X » « of X » « rate of X ¼ ¬ ¼ ¬ ¼ ¬ ¬ ¼. This is the basis of all mass balance calculations. It is built on the theory of mass conservation: matter is neither created, nor destroyed. The discovery of this, in the late 17th century, is credited to the French alchemist Antoine Laurent Lavoisier. The equation is valid for most chemical reactions, although it was amended by the more complete law of conservation of massenergy (one of the most famous equations in the world, a.k.a. E=mc2), formulated by Albert Einstein in 1905 (see Bodanis, 2000). If mass balance is applied to an aquatic system and complete mixing is assumed, this equation is more appropriately written as: V. wC wt. Qin Cin Qout Cout KT u V u C. (2). where V is the water volume, C is the concentration of the given substance in the water, Q is the water flow, KT is the turnover rate. In and out denote properties for inflowing and outflowing water, respectively. By replacing the differential equation with a difference equation and discretisation of the variables, modern computers can solve complicated systems of mass balance equations by using numerical methods (see Robertson, 1991; Goldstine, 1972).. 11.

(188) In aquatic foodweb modelling, mass is also conserved and eq. 1 can be applied. A principle setup for one primary unit (e.g., prey fish), PU, is given by: BM PU t

(189) BM PU t dt

(190) IPRPU CON PUSU ELPU

(191) u dt. (3). where BM is the biomass, IPR the initial production rate, CON the consumption rate (SU=secondary unit) and EL the elimination rate (e.g., by death or fishing; Håkanson and Boulion, 2002). Since 1968, when Vollenweider presented the first useful mass balance model for phosphorus in lakes (Vollenweider, 1968; 1969), there has been a plethora of eutrophication models for lakes (e.g., Jørgensen, 1976; Lung et al., 1976; Schnoor and O’Connor, 1980; Voivinov and Svirezhev, 1984; Del Furia et al., 1995; Malmaeus and Håkanson, 2004). A majority of them are based on mass balance equations for nutrients. Paper V discusses and reviews many studies of the effect of aquaculture on aquatic ecosystems, both for freshwater and marine systems and at different temporal and spatial scales.. 2.1.2 Fish farming and the mesocosm model A lot of work has been presented on the environmental impact of nutrient emissions from fish farms at various temporal and spatial scales; on the local scale by, e.g., Stigebrandt et al. (2004), on the regional scale by, e.g., Nordvarg and Johansson (2002) and on the Baltic scale by, e.g., Enell (1995). The mesocosm model (MESOMOD) is constructed to specifically model the nutrient dynamics to, from and within a mesocosm fish farm (Paper I). Marine fish farms are often located in coastal areas. Since both nitrogen and phosphorus can be the limiting nutrient for coastal areas (Kirkkala et al., 1998), the model was constructed to simulate the dynamics of both these nutrients. The basic flows for nitrogen can be seen in figure 1.. 12.

(192) Figure 1. Nitrogen flows in a fish farming mesocosm.. The model can be used as a basis for studying nutrient dynamics for different feeding strategies.. 2.1.3 The concept of functional groups At first sight, it might appear that the best ecological model is the one that accounts for a complete setup of processes. However, a very detailed model is no better than its input data and many environmental variables are difficult to measure and evaluate. As an example, the model for the benthic subsystem of the Ems estuary model (Admiraal et al., 1988) was very detailed and included aerobic and anaerobic bacteria which alone needed 30 parameters to be modelled. Evidently, if only few measurements are available, this is not a very ’user friendly’ model type. The inclusion of many variables that are difficult to measure leads to propagating uncertainties through the model which affects the overall model uncertainty. Obviuosly, a very simplified model also has weaknesses, e.g., the Vollenweider model, which does not give seasonal variations. This leads to the search of an optimal size for a model (see Håkanson, 1995). A key issue is to find ’collective’ parameters, which imply that low uncertainties can arise due to mutual compensations of processes occurring in the aquatic system (Monte, 1996). The use of functional groups for predictive modelling of aquatic ecosystems is a way of accomplishing this (Paper IV). The effect of physiological detail on model performance in marine ecosystem models has also been studied by Fulton et al. (2004).. 13.

(193) A functional group is a ’collective’ variable, where one does not look at the species level, but rather the niche level. The functional groups in the lake foodweb model LakeWeb (Håkanson and Boulion, 2002) are shown in figure 2.. Figure 2. The nine functional groups in the LakeWeb aquatic foodweb model.. 2.1.4 From LakeWeb to BaltWeb In order to transform a lake foodweb model into a marine foodweb model there are obviously several adjustments that have to be made. Paper IV is a step towards this transition, but the work should be regarded as preliminary and not the presentation of a final marine foodweb model. However, the relevance and scientific potential of the predictions that BaltWeb does, justify a presentation of this preliminary study. In order to make the model more reliable, access is needed to more data on habitat requirements, food choices, abiotic/biotic interactions and positive and negative feedbacks for the marine biomasses. The main reason why we present preliminary results for single fish species is that a high percentage of the predatory fish in the Baltic is cod (Gadus morhua) and a high fraction of the prey fish is herring (Clupea harengus) and sprat (Sprattus sprattus). Hence, the transformation of the LakeWeb model into BaltWeb is more straightforward.. 2.1.5 Fish farming as a point source of nutrients The environmental effects of fish farming are related to overuse of biotics, escaping fish, spreading of diseases to and causing genetic drift on natural populations (e.g., Hutchinson, 1997; Dar, 1999; FAO, 2001; Youngson et al., 2001; Read and Fernandez, 2003). However, in this thesis, the study is restricted to the impact of nutrient emissions of marine fish farming in the Baltic Sea (Papers I, II and V).. 14.

(194) 2.2 Image segmentation Image segmentation for identifying natural units in spatial datasets is a rapidly developing field of image processing. In its early years, it was a narrow research field focused on zone design (Openshaw; 1977;1978), but it has developed rapidly (Martin, 2000) partly due to the access to faster computers (Openshaw, 1994). Recently, various techniques for image segmentation, delineation and map regionalisation have been used (e.g., Cheng and Li, 2003; van der Sande et al., 2003; Kang et al., 2004; Luo et al., 2004). Segmentation is of interest in a variety of contexts (e.g., medicine, photography, remote sensing and GIS) where computational compartments are needed for descriptive or analytical purposes. Compartments are usually either defined according to homogeneity or by naturally occurring boundaries. Aquatic models can handle spatiality in different ways. The methods to divide the geographic model domain are either objective or subjective. Subjective methods include arbitrary defined units, where the user determines basin boundaries using the personal overall knowledge of the system. The success of these methods varies with the geographical complexity of the domain and/or the user’s skills. Models that use subjective methods are, e.g., Gieske (1997). Objective ways are commonly restricted to the application of a grid with a fixed (e.g., Lin et al., 2000), nested (e.g., Helminen et al., 1998) or variable (e.g., Burchard and Beckers, 2004) resolution. This creates a rasterized waterscape. One of the main problems in this context is to find an optimal spatial resolution. A coarse resolution grid obviously gives rise to crude simplifications of spatial properties, while a fine resolution requires input data with higher resolution and creates propagating uncertainties in the model. Evidently, basin delineation is not always problematic. In coastal bays with clearly defined boundaries (such as fjords) or in sea areas with distinct waterscape basins, there is no need for an objective method. In those cases, the modeller easily sets the boundaries. However, when moving to intermittent scales, e.g., river deltas or archipelago areas, where no clearly defined boundaries or subbasins can be identified (see Fig. 3 and Fig. 4), an objective method for this process would clearly be beneficial. The method presented in Paper IV is the first automated method developed for solving this problem. For areas with an organised drainage system (e.g., a stream network), there are several existing modelling methods to define subbasins based on digital elevation models (DEM) (see Goodchild et al., 1993). In contrast to existing delineation methods, the work presented in Paper IV uses trough filling algorithms combined with upstream cost flows to identify semi-enclosed com15.

(195) partments in topographically complex waterscapes without an organised drainage system. Since it uses a DEM, the delineation method is threedimensional but the resulting subbasin can be used in two-dimensional modelling.. 16.

(196) 3 Regional scale study areas. For the construction and testing of the WASUBI method (Paper III), two very different regions were selected. Although situated in different parts of the world and with completely different flora and fauna, the Finnish Archipelago Sea and the Botswanan Okavango Delta are related by their waterscape complexity (Fig. 3 and Fig. 4). They are both characterized by complex hydrological flow patterns.. 3.1 The Archipelago Sea, Finland The Finnish Archipelago Sea is one of the most island-rich archipelagos in the world, comprising more than 25,000 islands (Fig. 3). It is situated off the southwest coast of Finland with an average depth of only 23 m and the deepest trench reaches 146 m. The water volume is 210 km3 and the total area is 9,400 km2. The total drainage area covers 9,000 km2 with strong anthropogenic influence (HELCOM, 1996, 1998). The area has been subjected to many eutrophication studies (e.g., Kirkkala, 1998; Kirkkala et al., 1998; Hänninen et al., 2000) and a sometimes fierce debate on the role of aquaculture on eutrophication and algal blooms (see Peuhkuri, 2002; Bruun et al., 2002). On the Baltic scale, the Archipelago Sea acts like a filter, capturing nutrients and pollutants that settle within the area and cause eutrophication and anoxic conditions (Larsson et al., 1985; Rosenberg et al., 1990). The Archipelago Sea digital elevation model (DEM) used in Paper IV was built with data from FMA (2000) and Seifert et al. (2001). The grid resolution was set to 100 m.. 3.2 The Okavango Delta, Botswana In northern Botswana, Southern Africa, lies one of the world’s largest inland deltas, the Okavango Delta (40,000 km2). The world’s attention was drawn to this region by the adventurer and explorer David Livingstone, who visited Lake Ngami on the southern fringe of the Delta in 1849 (Livingstone, 1857). 17.

(197) Several famous travellers followed during the early African colonial period, but the Delta remained mainly a geographical curiosity, mostly due to its remoteness, the threats posed by wild animals and frequent occurrence of dangerous diseases, including malaria and sleeping sickness. The Okavango River flows from the humid tropical highlands of central Angola and stretches out into the Kalahari Desert. Here, almost all of the water inflow (about 10 km3/year) is lost due to evapotranspiration and only about 2% of the water input (river inflow plus rainfall) annually leaves the system as surface flow (McCarthy and Ellery, 1994). The Delta (Fig. 4) is a large alluvial fan situated in a fault-bounded depression and contains both permanent and seasonal swamps. The Delta surface is extremely flat and the average slope of the fan is 0.00035 from its apex to its base. (Gumbricht et al., 2002). Water depth usually does not exceed 2 m except in some channels and lagoons (Dincer et al., 1987). The waters sustain about 4,000 km2 of permanent swamps that expand to up to 12,000 km2 in years with plentiful precipitation (McCarthy et al., 2003). The Okavango is one of the few remaining large river regions in the world without manmade developments, but as the demand for the limited fresh water grows, the sensitive ecology of the Delta might face large changes. The international conflict, due to Namibia’s plans to extract water from the river, is now likely to escalate, as Angola plans to develop the area after the end of a two-decade long civil war. This is part of the reason for the increasing interest in the Okavango Delta hydrological system (Scudder et al., 1993) and in hydrological modelling of the Delta (Dincer et al., 1987; Gieske, 1997; Andersson et al., 2003; Bauer et al., 2004). For our delineation method (Paper III), a DEM from Gumbricht et al. (2004) was utilized. Using landcover and islands to infer local relief, a relative microtopographical map with 28.5 m resolution over the Okavango Delta has been constructed. For computational reasons, the DEM used in Paper IV was resampled to 500 m.. 18.

(198) Figure 3. The Archipelago Sea. A. LANDSAT TM composite. B. DEM. C. Aerial photo (Photo: Unknown)..

(199) Figure 4. The Okavango Delta. A. Landsat TM composite. B. DEM. C. Aerial photo (Photo: T. Gumbricht).

(200) 4 Highlighted results and discussions. 4.1 The mesocosm nutrient model (Paper I) With the mesocosm model (Paper I) the nutrient fluxes were quantified. According to table 1, nitrogen fluxes were largest in the outflowing water. For phosphorus, the uptake in rainbow trout approximately equalled the outflow. Almost 26% of the phosphorus sedimented during the experiment and the uptake of nutrients in mesocosm biomass was about 7% of the total nutrient output. The model was found to be sensitive for the growth pattern of the rainbow trout although none of the six tested growth submodels could be easily rejected. Model performance for the chosen growth submodel, based on an empirical metabolic growth equation moderated by feed intake, can be seen in figure 5.. Figure 5. Modelled values versus empirical data for the calibrated mesocosm model. 95% confidence bands are shown as dashed lines.. 21.

(201) Table 1. Simulated mesocosm nutrient flows (total flow, HER-feed scenario). Nutrient flows. Phosphorus. Nitrogen. % of total. % of total. Input Incoming water Farm –> water TOTAL. 22.9 77.1 100. 30.7 69.3 100. Output Water outflow Water –> fish Water –> sediment Sediment –> atmosphere Water –> atmosphere Water –> biomass TOTAL. 33.9 32.5 25.9 0 0 7.8 100. 54.3 23.8 7.2 6.6 2.0 6.1 100. One of the aims of the study was to calculate the sustainability threshold value for a fish farm when locally caught herring or sprat was included in the fish feed. The results of the threshold simulations showed that an inclusion of 11% was sufficient for reaching this goal (Fig. 6). In this context, sustainability is defined as aquaculture having a zero net nutrient load on a regional scale.. 60 Herring inclusion. 0%. 20. 10% -20. 20% 30%. -60. 45%. -100. 60% -140 94-06-23. 94-08-02. 94-09-11. 94-10-21. date. Figure 6. Simulations using different herring inclusions in fish feed.. 22.

(202) 4.2 The SPM model for coastal areas (Paper II) One working hypothesis of this work was that there may be major differences in the ranking of the fluxes of suspended particulate matter (SPM) between coastal areas close to the Sea and less exposed coastal areas deep in archipelago areas. However, the results show that only minor differences between the areas can be identified (Fig. 7). The inflow of SPM to the surface water is the dominating flux in all areas. A difference can be seen for SPM flux due to land uplift, since the coastal area Guavik is situated in southern Sweden where land uplift is zero. It is interesting to note that the fluxes from primary production are relatively small in all three coastal areas. This means that the conditions in the Sea play an important role for the transport of SPM and hence for all pollutants associated with carrier particles included in this group in most or maybe all Baltic coastal areas. The main reason for this is that the characteristic theoretical retention time of surface water is short for these coastal areas. 100000. Annual flux (tonnes/year). L. Rimmoe Guavik. 10000. Laitsalmi. 1000. 100. 10. Farm to DW. Inflow to DW. Min DW. Farm to SW. Prim prod. Burial. Min SW. Mix DW to SW. Outflow from DW. Sed to DW. Res to DW. Land uplift. Res to SW. Mix SW to DW. Sed on A. Sed on ET. Inflow to SW. Outflow from SW. 1. Figure 7. SPM fluxes for different processes in the dynamic model (Model presentation and abbreviations are given in Paper II). L. Rimmö is in direct contact with the Baltic Sea; Guavik is exposed to the Baltic Sea and Laitsalmi is deep in the Finnish archipelago.. 23.

(203) 4.3 The delineation of topographically complex waterscapes (Paper III) A waterscape is a sub watersurface landscape. For a waterscape with large topographical complexity, the morphometric properties are decisive for the hydrodynamics of the aquatic system. The method presented in paper III is an attempt to use the information in digital elevation models (DEMs) to regionalize such a waterscape. The WASUBI (WAterscape SUBbasin Identification) method is designed to be run with only a DEM as input data and to be very user friendly. The single demand on the user is to give WASUBI the desired number of final subbasins. The basic workflow of the WASUBI method (Paper III) is given in figure 8. After the identification and filling of all local troughs, all troughs are sorted according to size and the largest ones (corresponding to the userdefined number) are selected as nuclei areas for further expansion into subbasins. The expansion process was tested with different cost growth functions that were shown to give slightly different results for delineations of regions with many sounds and straights (Paper III). Figure 9 shows the delineation results for the Archipelago Sea and the Okavango Delta. Tests showed that the method was sensitive to the resolution of the DEM. This is not unexpected, as different delineations will be produced when the topographic complexity is simplified. When a coarser grid is created, sounds and straights that caused a high cost at a fine resolution simply merge with land areas and disappear due to the interpolation process. With a simple test (see Paper III), the openness between the WASUBI subbasins was shown to be lesser than between semi-randomly created subbasins (table 2). The area and volume of the WASUBI basins also show a much larger variance indicating that the method works in an intended way, since size is of minor relevance when it comes to openness and homogeneity. A visual inspection of both study areas reveals that the largest subbasins are likely to have a faster internal turnover and mixing compared to the exchange with adjacent basins.. 24.

(204) Figure 8. Basic workflow of the WASUBI delineation method.. Figure 9. The WASUBI delineation of the study areas (The Archipelago Sea and the Okavango Delta) into 15 subbasins.. 25.

(205) Table 2. Comparison between 14 subbasins generated over the Okavango by WASUBI and with a semi-random method (see text). WASUBI Parameter Boundary length (m) Area (km2) Volume (m3*106) Openness (m2). Random. Mean. SD. Mean. SD. 55065 926 510 11081. 59935 1187 1260 19749. 65798 927 511 38498. 45840 644 689 48700. 4.4 The BaltWeb simulations (Paper IV) The BaltWeb model was used to simulate scenarios of the predatory cod stock biomass dynamics for different fishing scenarios (Fig. 10). In the default setup, the model was run with a fishing rate of 75 ktons/year of predatory cod and at the start of the simulations, a fishing of 110 ktons/year was assumed. This is equivalent to today’s allowed catch of 70 ktons/year plus an estimated illegal catch of 40 ktons/year. The results from the different scenarios in figure 10 show the effects of a return to a cod fishing rate of 75 ktons/year after x years. Under these presuppositions it is interesting to note, that the cod stock will not recover within ten years if the overfishing would allow the stock biomass to go below 120 ktons.. Figure 10. Simulations of fish biomasses from different cod fishing scenarios.. 26.

(206) 4.5 Environmental consequence analyses of fish farm emissions (Paper V) This paper is a review and fish farming is used as a case study of nutrient point source emissions. The basic idea is to put these emissions into a context at various spatial scales. With examples for the site, local, regional and international scales, measurements and modelling are combined to quantify nutrient fluxes and relate them to fish farm emissions. An overview of the available literature on this matter shows that aqutic modelling is less common on the intermittent scales. Therefore it is hypothesized that aquatic modelling on the regional scale (e.g., several adjacent coastal areas or an archipelago) can benefit from new GIS techniques for spatial modelling. Mass balance modelling can be done in many different ways (different scales, target variables, driving variables and mathematical structures) and criteria for the most optimal scale of practical aquatic management are discussed.. 27.

(207) 5 Main findings and conclusions. x The mesocosm model (Paper I) has been shown to produce good predictions of the nutrient dynamics at this scale. The growth pattern of the cultivated fish affects the nutrient concentrations in the mesocosm water to a large extent. x In order to have a zero net nutrient load from the fish farm, the mesocosm model shows that it is sufficient to use an 11% inclusion of regionally caught wild fish in the fish feed. x By quantification and ranking of SPM fluxes to, within and from coastal areas with different morphometric characteristics and distances to the Sea (Paper III), it has been shown that the conditions in the Sea are of fundamental importance, also for the most enclosed and sheltered coastal areas. This is relevant for an understanding of pollution transport in coastal areas and to get realistic expectations of remedial actions at local, regional and international scales. x The concept of scale is discussed in Paper V. This is a very important concept in aquatic modelling. With knowledge of the importance of uncertainties, data availability and the relevance of different fluxes at various scales, aquatic mass balance modelling can be optimized. x At an international scale, the modelling focus has to be set on major fluxes (Paper V). With the BaltWeb approach (Paper IV), an attempt is made to explore a new way of setting fish quotas based on environmental key factors. If it is further developed and validated it could be an important complement to methods available and used today. x At the intermittent regional scales, morphometry is found to be an important regulating factor (paper II) for aquatic ecosystem modelling. x The WASUBI delineation method (Paper III) is meant to be a userfriendly and practically useful method for objective subdivision of topographically complex waterscapes. The method is shown to delineate enclosed subbasins. However, WASUBI is sensitive to the DEM resolution. 28.

(208) and further testing is needed to validate or improve the subbasin nuclei selection process. x Recognizing morphometry as a key factor for aquatic modelling at intermittent scales the WASUBI delineation method could be used to improve aquatic ecosystem modelling at a regional scale.. 29.

(209) 6 Acknowledgements. There are many people that in one way or another have contributed to this work and have followed me on this long journey. It would not have been joyful, or even possible, to finish this work alone, and to those that have helped me forward, I owe many thanks. I express my deepest gratitude to… … Professor Lars Håkanson, my supervisor, mentor and discussion partner. Without his enthusiasm, focus and great belief in new ideas I would never have started as a PhD student. We have had many interesting discussions on holistic views of ecosystem modelling. … Thomas Gumbricht, my dear friend and partner in GIS and field research. I owe you many thanks for giving me the opportunity to discover Africa. Under sometimes rough conditions, we have combined doing science with great wildlife experiences. … Thorsten Blenckner, my new friend, colleage and coach. You were of uttermost importance at critical moments in the finalisation of this work. … my friends from the start of my aquatic and environmental engineering studies: Niclas Bockgård, Klas Hansson, Fredrik Wetterhall, Johan Öhman and Mikael Malmaeus. For almost ten years, you have been a part of my everyday environment. Thanks for sharing laughs and your thoughts on just about everything (including science!). … my present and former colleagues: Angelica, Aziz, Tony, Håkan, Jonas, Shulan, Karin, Andreas, Sara-Sofia, Elin, Karin, Torbjörn, Lennart, LarsChrister, Allan, Conny and everyone else at the LUVA research group. … Per Jonsson for letting me escape my office once in a while and sign on for “sediment destroyer” R/V Sunbeam. Either as student, field course assistant, crew or friend. I also owe thanks to Johan Persson, crew colleague and my first roommate at Geocentrum. … Tomas Nord, for interesting discussions and short relaxing email communications. We share the same peculiar interest for the computer industry, although we are definitely not on the same team. … Staffan Holmgren, for introducing me to the wonderful world of aquatic sciences, about 30 years ago. With you and your family, we have not only studied nature, but also enjoyed it. … Jan-Erik Åslund, for showing faith in me and my skills, although I was young, by giving me the opportunity to work with environmental monitoring in Jämtland. 30.

(210) … Riddarhuset, for financial support on scientific travels and conferences. … Apple Computer, for providing a bicyle for my brain. My Macs and my iPod always help me out and remind me to think different. … Joakim Eriksson, my dear friend. You’re always there for me, and we share the same interest for many things in life. I look forward to many more years of friendship, travels and interesting conversations together. … my musical friends in Uppsala Blåsarsinfonietta, Velodrom and Chorus Virorum, for making life in Uppsala so much more than work. … my late grandfather Per Lönn, for helping me become who I am today. Thanks for Sundsjön, the place in my heart! … Oskar, my dear cousin, saxophone partner and friend. Having you around is always a joy and a comfort. … my parents, Roger and Monica, for always believing in me and supporting me, whatever I’ve tried to do or participate in. That has definitely meant a lot to me. … my brother Gustav and my sister Hanna, for being the best possible brother and sister one can have and for always being there when its needed. Finally, Solveig. Without you, where would all my inspiration and happiness be? Not to say my English! Your smile and your love carry me through everything. Thank you for your endless care and support. You are the sunshine of my life.. 31.

(211) 7 Summary in Swedish. Denna avhandling presenterar massbalansmodeller för akvatiska ekosystem. Geografiska informationssystem (GIS) utgör en viktig del av avhandlingen. Massbalansmodellerna fokuserar på flöden av närsalter och suspenderat partikulärt material, biotiska/abiotiska samband och rör olika temporala och spatiella skalor. Skalperspektivet och dess roll inom massbalansmodellering diskuteras speciellt. I syfte att studera och uppskatta närsaltsbelastningen från fiskodlingar konstruerades en modell baserat på ett mesokosmförsök. Modellen simulerar flöden av fosfor och kväve från, till och inom en kassodling för regnbågslax (Oncorhynchus mykiss). En känslighetsanalys av modellen visar att de största osäkerheterna finns i regnbågslaxens tillväxtmönster. Olika delmodeller för fisktillväxt, antingen befintliga eller utvecklade i denna studie, testades. Modellen användes till att upskatta hur stor inblandning av regionalt fångad fisk som krävs för att åstadkomma en miljömässigt hållbar fiskodling. Hållbarhet definieras i detta sammanhang som då nettotillskottet av näringsämnen från fiskodlingen på en regional skala är noll. Resultatet visar att en inblandning på 11% av vildfisk i den odlade fiskens foder räcker för att uppnå målet. I ett skalperspektiv ses fiskodlingen som en punktkälla men som sätts in i ett regionalt perspektiv. En uppsats beskriver en modell för suspenderat partikulärt material (SPM) och sedimentation inom kustområden (den lokala skalan). Modellen appliceras på kustområden i Östersjön med vitt skilda egenskaper beträffande morfometri och avstånd till öppet hav. Fiskodlingar används här igen som exempel på punktkälla för näringsutsläpp och effekterna av olika stora fiskodlingar i olika kustområden modelleras. Resultaten visar att förhållandena i öppna havet har mycket stor betydelse för alla undersökta kustområden, även de mest inneslutna. Detta beror på att vattnets teoretiska utbytestid inom dessa områden är kort. I denna studie förs skalresonemanget över till ett mer regionalt och även internationellt perspektiv. Med utgångspunkt från en vältestad näringsvävsmodell för sjöar, LakeWeb, utforskas möjligheterna till ett nytt förhållningssätt beträffande förutsättningarna som reglerar vilka fiskekvoter som kan/bör sättas för Östersjön. Näringsväven i LakeWeb bygger på funktionella nyckelarter istället för på 32.

(212) enskilda arter. Detta gör den principellt olämplig för att modellera exempelvis enskilda fiskbestånd men eftersom Östersjön är relativt artfattig kan ansatsen ändå rättfärdigas. Modellen transformeras till och kalibreras för förhållandena i egentliga Östersjön. LakeWeb innehåller även en massbalansmodell för fosfor som tar hänsyn till viktiga miljöfaktorer (som salinitet, näringsstatus, temperatur och sedimentation). För att visa ansatsens potential presenteras simuleringar för olika fiskescenarier. Effekten av illegalt fiske simuleras och även hur ett reducerat torskfiske påverkar biomassorna av strömming och skarpsill. Med denna typ av modellering kan alltså mycket intressanta frågeställningar beträffande näringsdynamiken i Östersjön och effekterna av olika åtgärder för fiskebestånden modelleras. Modellen kräver dock noggrann kalibrering och validering innan mer tillförlitliga slutsatser kan dras för fisket i Östersjön. Den mest komplicerade skalan för akvatiska massbalansmodeller är den regionala. För att förbättra dynamisk massbalansmodellering på denna skala har ett program tagits fram för indelning av topografiskt komplexa vattenområden, exemelvis deltan och skärgårdar, där de hydrodynamiska förhållandena är svårmodellerade. Programmet, WASUBI, är användarvänligt och kräver minimalt med indata. Med utgångspunkt från en digital höjdmodell över området skapas ett, av användaren bestämt, antal delbassänger. Metoden testades för två skilda områden, det finska Skärgårdshavet samt Okavangodeltat i Botswana. Utifrån kärnområden, skapade med en algoritm som fyller upp lokala sänkor i datasetet, tillåts dessa växa ihop med hjälp av en friktionsyta. Tre olika typer av friktionsytor testades och dessa gav bara upphov till små skillnader i bassängindelningen. Metoden skapar delbassänger med större inneslutenhet än en slumpmässig indelningsmetod. Tester visar att modellen är känslig för den rumsliga upplösningen på höjddata. Detta har sin förklaring i att trösklar och sund försvinner i och med interpolationen till ett dataset med grövre upplösning. Sammanfattningsvis diskuteras näringutsläpp på olika skalor i Östersjön och en litteraturstudie samlar den tillgängliga kunskapen om fiskodlingars påverkan på platsen för odlingen, den lokala, regionala samt internationella skalan. Om man tar hänsyn till de olika förutsättningar, exempelvis beträffande målvariabler och viktiga processer, som råder på de olika skalorna kan akvatisk massbalansmodellering förbättras jämfört med idag. Kan man dessutom områdesindela den svårmodellerade regionala skalan med hjälp av objektiva GIS-metoder är förutsättningarna goda för att uppnå ökad prediktionskraft för akvatiska ekosystemmodeller.. 33.

(213) 8 References. Admiraal, W., van Arkel, M.A., Baretta, J.W., Colijn, F., Ebenhöh, W., de Jonge, V.N., Kop, A., Ruardij, P. and Schröder, H.G.J., 1988. The construction of the benthic submodel. In: Baretta, J.W. and Ruardij, P. Tidal flat estuaries. Simulation and analysis of the Ems estuary. Ecol. Studies 71: 105-152. Andersson, L., Gumbricht, T., Hughes, D., Kniveton, D., Ringrose, S., Savenije, H., Todd, M., Wilk, J. and Wolski, P., 2003. Physics and Chemistry of the Earth, 28: 1165-1172. Bauer, P., Thabeng, G., Stauffer, F. and Kinzelbach, W., 2004. Estimations of the evapotranspiration rate from diurnal groundwater level fluctuations in the Okavango Delta, Botswana. Journal of Hydrology, in press. Bodanis, D., 2000. E=mc2 : a biography of the world's most famous equation. Macmillan, London. Bruun, H., Hukkinen, J. and Eklund, E., 2002. Scenarios for coping with contingency: The case of aquaculture in the Finnish Archipelago Sea. Technological Forecasting & Social Change, 69: 107-127. Burchard, H. and Beckers, J.-M., 2004. Non-uniform adaptive vertical grids in onedimensional numerical ocean models. Ocean Modelling, 6: 51-81. Del Furia, L., Rizzoli, A. and Arditi, R., 1995. Lakemaker: A general object-oriented software tool for modelling the eutrophication process in lakes. Environmental Software, 10: 43-64. Cheng, H.D. and Li, J., 2003. Fuzzy homogeneity and scale-space approach to color image segmentation. Pattern Recognition, 36: 1545–1562. Dar, W., 1999. Sustainable aquaculture development and the Code of Conduct for Responsible Fisheries. http://www.fao.org/fi/meetings/minist/1999/dar.asp Dincer, T., Child, S. and Khupe, B., 1987. A simple mathematical model of a complex hydrologic system – Okavango Swamp, Botswana. Journal of Hydrology, 93: 41-65. Enell, M., 1995. Environmental impact of nutrients from Nordic fish farming. Water Science and Technology. 31: 61-71. FAO, 2001. Report of the Conference on Aquaculture in the Third Millennium. FAO Fisheries Report No. 661, FIRI/R661(En). ISBN 92-5-104708-1. FMA (Finnish Maritime Administration), 2000. FINGIS digital maritime charts corresponding to Finnish National Land Survey 1:50 000 base maps. Fulton, E.A., Parslow, J.S., Smith, A.D.M. and Johnson, C.R., 2004. Biogeochemical marine ecosystem models II: the effect of physiological detail on model performance. Ecological Modelling, in press. Gieske, A., 1997. Modelling the outflow from the Jao-Boro river system in the Okavango delta, Botswana. Journal of Hydrology, 193: 214-239. Gumbricht, T., McCarthy T.S. and Merry C., 2002. Topography of the Okavango Delta, Botswana, and sedimentological and tectonic interpretations. South African Journal of Geology, 104: 243-264.. 34.

(214) Gumbricht, T., McCarthy T.S. and Bauer P., 2004. Microtopographic relief of the Okavango Delta Botswana. Accepted for publication in Earth surface processes and landforms. Goldstine, H.H., 1972. A history of the calculus of variations from the 17th through the 19th century, Princeton U.P. Goodchild, M.F., Parks, B.O., Steyaert, L.T., 1993. Environmental modelling with GIS. Oxford University Press. Håkanson, L., 1995. Optimal size of predictive models. Ecological Modelling, 78: 195-204. Håkanson, L. and Boulion, V., 2002. The Lake Foodweb – modelling predation and abiotic/biotic interactions. Backhuys Publishers, Leiden. HELCOM, 1996. Third periodic assessment of the state of the marine environment of the Baltic Sea, 1989-1993; Background document. Baltic Sea Environmental Proceedings 64B, 1-252. HELCOM, 1998. The Third Baltic Sea Pollution Load Compilation (PLC 3) Balt. Sea Environ. Proc. No. 70. Helminen, H., Juntura, E., Koponen, J., Laihonen, P. and Ylinen, H., 1998. Assessing of long-distance background nutrient loading to the Archipelago Sea, northern Baltic, with a hydrodynamic model. Environmental Modelling & Software, 13: 511-518. Hutchinson, P. (ed), 1997. Interactions between Salmon Culture and Wild Stocks of Atlantic Solmon: The Scientific and Management Issues. ICES Journal of Marine Science, 54: 963-1227. Hänninen, J., Vuorinen, I., Helminen, H., Kirkkala, T. and Lehtilä, K., 2000. Trends and Gradients in Nutrient Concentrations and Loading in the Archipelago Sea, Northern Baltic, in 1990-1997. Estuarine, Coastal and Shelf Science, 50:153171. Jørgensen, S.E., 1976. A eutrophication model for a lake. Ecological Modelling, 2: 147-165. Kang, Y., Engelke, K.and Kalender, W.A., 2004. Interactive 3D editing tools for image segmentation. Medical Image Analysis, 8: 35-46. Kirkkala, T., 1998. Miten voit Saaristomeri? Lounais-Suomen ympäristökeskus. Ympäristön tila Lounais-Suomessa 1 (in Finnish). Kirkkala, T., Helminen, H. and Erkkilä, A., 1998. Variability of nutrient limitation in the Archipelago Sea, SW Finland. Hydrobiologia, 363: 117-126. Lin, B.-L., Sakoda, A., Shibasaki, R., Goto, N. and Suzuki, M., 2000. Modelling a global biogeochemical nitrogen cycle in terrestrial ecosystems. Ecological Modelling, 135: 89-110. Lung, W.S., Canale, R.P. and Freedman, P.L., 1976. Phosphorus models for eutrophic lakes. Water Research, 10: 1101-1114. Livingstone, D., 1857. Missionary travels and researches in South Africa. John Murray, London. Luo, J., Singhal, A., Etz, S.P. and Gray, R.T., 2004. A computational approach to determination of main subject regions in photographic images. Image and Vision Computing, 22: 227–241. Malmaeus, M. and Håkanson, L., 2004. Development of a Lake Eutrophication model. Ecological Modelling, 171: 35-63. Martin, D., 2000. Automated zone design in GIS, in "Innovations in GIS 7". Taylor and Francis, London. McCarthy, T.S. and Ellery, W.N., 1994. The effect of vegetation on soil and ground water chemistry and hydrology of islands in the seasonal swamps of the Okavango Fan, Botswana. Biological Conservation, 70: 159-168.. 35.

(215) McCarthy, J., Gumbricht T., McCarthy, T.S., Frost P., Wessels K. and Seidel, F., 2003. Flooding dynamics of the Okavango wetland in Botswana between 1972 and 2000. Ambio, 32: 453-457. Monte, L., 1996. Collective models in environmental sciencies. The Science of the Total Environment, 192: 41-47. Mäkinen, T., 1991. Utsläpp av kväve (N) och fosfor (P) samt organisk stof från havbruket. In: Havbrug og miljø, Torshavn, Faroe Island, Sept 12-14 1990. Hoffmann, E. (red.). Nordic Council of Ministers. ISBN 91-7996-318-8. Openshaw, S., 1977, Optimal zoning systems for spatial interaction models. Environment and Planning A, 9: 169-184. Openshaw, S., 1978. An empirical study of some zone design criteria, Environment and Planning A, 10: 781-794. Openshaw, S., 1994. A concepts-rich approach to spatial analysis, theory generation, and scientific discovery in GIS using massively parallel computing, in M. Warboys (ed.), Innovations in GIS 1, Taylor and Francis, London. Peuhkuri, T., 2002. Knowledge and interpretation in environmental conflict Fish farming and eutrophication in the Archipelago Sea, SW Finland. Landscape and Urban Planning, 61: 157-168. Robertson, J.S., 1991. Numerical methods, in Simmons, G.F., Differential equations with applications and historical notes, McGraw-Hill, Singapore. Read, P. and Fernandes, T., 2003. Management of environmental impacts of marine aquaculture in Europe. Aquaculture, 226: 139-163. van der Sande, C.J., de Jong, S.M. and de Roo, A.P.J., 2003. A segmentation and classification approach of IKONOS-2 imagery for land cover mapping to assist flood risk and flood damage assessment. International Journal of Applied Earth Observation and Geoinformation, 4: 217-229. Scudder, T., Manley, R.E., Coley, R.W., Davis, R.K., Green, J., Howard, G.W., Lawry, S.W., Martz, D., Rogers, P.P., Taylor, A.R.D., Turner, S.D., White, G.F., Wright, E.P., 1993. The IUCN Review of the Southern Okavango Integrated Water Development Project. IUCN, Gland, Switzerland. 544 pp. Schnoor, J.L. and O’Connor, D.J., 1980. A steady state eutrophication model for lakes. Water Research, 14: 1651-1665. Seifert, T., Tauber, F., Kayser, B., 2001. A high resolution spherical grid topography of the Baltic Sea – revised edition. Proceedings of the Baltic Sea Science Congress, Stockholm 25-29. November 2001 (to be published). Stigebrandt, A., Aure, J., Ervik, A. and Hansen, P.K., 2004. Regulating the local environmental impact of intensive marine fish farming III. A model for estimation of the holding capacity in the Modelling–Ongrowing fish farm–Monitoring system. Aquaculture (in press). Subasinghe, R., 2004. Aquaculture in the global food supply. FAO, Rome. Voivinov, A.A. and Svirezhev, Yu.M., 1984. A minimal model of eutrophication in freshwater ecosystems. Ecological Modelling, 23: 277-292. Vollenweider, R.A., 1968. The scientific basis of lake eutrophication, with particular reference to phosphorus and nitrogen as eutrophication factors. Tech. Rep. DAS/DSI/68.27, OECD, Paris, 159pp. Vollenweider, R.A., 1969. Possibilities and limits of elementary models concerning the budget of substances in lakes. Archive für Hydrobiologie, 66: 1-36. Youngson, A.F., Dosdat, A., Saroglia, M. and Jordan, W.C., 2001. Genetic interactions between marine finfish species in European aquaculture and wild conspecifics. Journal of Applied Ichthyology, 17: 153-162.. 36.

(216) Appendix A. WASUBI source code. This appendix contains a compilation of all Avenue scripts in the WASUBI delineation method. All code is developed within the study presented in Paper III, except when explicitly stated otherwise. Table 3. WASUBI scripts Name. Short description. main.ave step1.ave step2.ave step3.ave step4.ave step5.ave step6.ave step7.ave step8.ave step9.ave step10.ave. 38 Runs as main program, connects to all other scripts. Creates a grid theme by filling all sinks in another grid theme. 40 Takes a filled depths theme and converts it to a grid. 42 Takes the grid theme and exports it to a shapefile. 43 Exports all the features in a theme in three steps. 45 Negates all depths in the shapefile for further use. 46 Exports the features according to depth to new themes. 47 Exports the features for nucleus areas and depths. 49 Takes the nucleus theme, sorts, exports and number the nucleus areas. 51 Takes the nucleus shapefile and splits it up into separate shape files. 53 Takes the list of nucleus shapefiles and creates cost distance grids for 54 all nucleus areas. Takes the list of cost distance grids and combines them to a least cost 55 distance grid with the closest nucleus number as grid code. Takes the subbasin grid theme and exports it to a shapefile. 56 Takes the filled DEM and calculates the areas where the dataset is 58 defined (i.e., removes outskirts). Creates shapefile mask, part I. 60 Creates shapefile mask, part II. 63 Takes the mask and clips the subbasin shapefile. 64 A cleaning script. 65. step11.ave step12.ave step13.ave step14.ave step15.ave step16.ave step17.ave. Page. 37.

(217) ' Name : main ' Version : 1.0 ' Author : Andreas Gyllenhammar ' Title : WASUBI ' -----------------------------------------------------------------------------' Description : Takes a topographically complex area and divides it ' : into a user defined number of subbasins. ' Runs : As main program ' Run by : ' Self : ' Returns : ' -----------------------------------------------------------------------------DEBUG = true '*** Create temporary library defbibl = av.GetProject.GetWorkDir.AsString comline = "c:\WINNT\system32\command.com /c mkdir " + defbibl + "\tempi" System.Execute(comline) nybibl = defbibl + "\tempi" av.GetProject.SetWorkDir(nybibl.AsFileName) '*** Choose View theProj = av.GetProject theDocList = theProj.GetDocs theDict = Dictionary.Make(theDocList.Count) for each i in theDocList if (i.is(View)) then theDict.Add(i.GetName,i) end end theList = theDict.ReturnKeys if (theList.Count = 1) then theView=theDict.Get(theList.Get(0)) else theView = theDict.Get(MsgBox.ListAsString(theList,"Choose a view","VIEWS")) if (theView=NIL) then exit end end '*** Choose GridTheme theThemes = theView.GetThemes theDict = Dictionary.Make(theThemes.Count) for each i in theThemes if (i.is(GTheme)) then theDict.Add(i.GetName,i) end end theList = theDict.ReturnKeys if (theList.Count = 1) then theTheme = theDict.Get(theList.Get(0)) else theTheme = theDict.Get(MsgBox.ListAsString(theList,"Choose a grid theme","GRID THEMES")) if (theTheme = nil) then exit end end '*** Let user set desired number of nucleus areas nucnum = msgbox.Input("Enter the desired number of nucleus areas","Nucleus areas","6").AsNumber '*** START THE WASUBI PROCESS starttid = Date.Now av.ShowMsg ("Creating a depressionless DEM") filleddepths = av.run( "step1" , { theView , theTheme } ) av.ShowMsg ("Converting to raster format") fillgrid = av.run( "step2" , { theView, filleddepths } ) av.ShowMsg ("Converting to vector format") initshapethm = av.run( "step3" , { theView, fillgrid } ) av.ShowMsg ("Finding nucleus areas") sortedthm = av.run( "step4", { theView , initshapethm } ) av.ShowMsg ("Negating data set"). 38.

(218) negpos = av.run ("step5", { theView , sortedthm } ) av.PurgeObjects av.GetProject.Save av.ShowMsg ("Exporting islands") islthm = av.run( "step6", { theView , negpos } ) av.ShowMsg ("Exporting nucleus areas") prenuclthm = av.run( "step7", { theView , negpos, nucnum } ) av.ShowMsg ("Sorting nucleus areas") nuclthm = av.run ( "step8", { theView , prenuclthm , negpos } ) av.ShowMsg ("Splitting nucleuses into separate grids") nuclthmlist = av.run ( "step9", { theView , nuclthm} ) av.PurgeObjects av.GetProject.Save av.ShowMsg ("Calculating cost distance") costdistthmlist = av.run ("step10", {theView, theTheme, nuclthmlist} ) av.ShowMsg ("Calculating total cost distance") subbasingrid = av.run("step11", { theView , costdistthmlist } ) av.ShowMsg ("Converting to vector format") unmaskedsubs = av.run("step12", {theView , subbasingrid } ) av.ShowMsg ("Creating mask, part I") mask1 = av.run("step13", {theView, fillgrid } ) av.ShowMsg ("Creating mask, part II") mask2 = av.run("step14", {theView, mask1 } ) av.PurgeObjects av.GetProject.Save av.ShowMsg ("Creating mask, part III") mask3 = av.run("step15", {theView, mask2 } ) av.ShowMsg ("Clipping sub basins") maskedsubs = av.run("step16", {theView, mask3, unmaskedsubs } ) av.PurgeObjects av.GetProject.Save '*** Clean up av.ShowMsg ("Cleaning up") clean = av.run("step17", {theView, maskedsubs, theTheme} ) Sluttid = Date.Now TT = Sluttid - Starttid time = TT.AsSeconds hours = (time/3600).Floor mins = ((time/3600-hours)*60).Floor secs = (time-hours*3600-mins*60).Floor rows = theTheme.GetGrid.GetNumRowsAndCols.Get(0) cols = theTheme.GetGrid.GetNumRowsAndCols.Get(1) MsgBox.Info( "Used time: " + hours.AsString + " hours " + mins.AsString + " minutes " + secs.AsString + " seconds." + nl + "Grid size: " + rows.AsString +" x " + cols.AsString + nl + "Time per pixel: " + (time / (rows * cols)).AsString + nl + nl + "WASUBI successful" , "The Flaskhals Process" ) System.Beep. 39.

(219) ' Name ' Author ' Title '. : step1 : Andreas Gyllenhammar : DEMFill (modified from Spatial.DEMFill) Creates a grid theme by filling all sinks in another grid theme. -----------------------------------------------------------------------------Description : Takes a grid theme, normalises it and fills all sinks, areas of internal drainage, contained within it. The aGrid.FlowDirection, aGrid.Sink, aGrid.Watershed, aGrid.ZonalFill, and aGrid.Con requests is used to fill the sinks. The process of filling sinks can create sinks, so a looping process is used until all sinks are filled. One cell sinks are not filled. Sinks of any depth are filled.. ' ' ' ' ' ' ' ' ' ' ' Runs : As step 1 ' Run by : main ' Self : The current view, the depth grid ' Returns : A depressionless depth grid ' -----------------------------------------------------------------------------DEBUG = true theView = SELF.Get(0) theTheme = SELF.Get(1) '*** Fill sinks in Grid until they are gone elevGrid = theTheme.GetGrid '*** Create normalised negative grid elevGrid = elevGrid.Float elevGrid = elevGrid.Abs elevGrid = elevGrid / (elevGrid.GetStatistics.Get(1)).AsGrid * 1000 elevGrid = elevGrid.Int elevGrid = elevGrid * -1.AsGrid sinkCount = 0 numSinks = 0 while (TRUE) flowDirGrid = elevGrid.FlowDirection(FALSE) sinkGrid = flowDirGrid.Sink if (sinkGrid.GetVTab = NIL) then 'Check for errors if (sinkGrid.HasError) then return NIL end sinkGrid.BuildVAT end 'Check for errors if (sinkGrid.HasError) then return NIL end if (sinkGrid.GetVTab <> NIL) then theVTab = sinkGrid.GetVTab numClass = theVTab.GetNumRecords newSinkCount = theVTab.ReturnValue(theVTab.FindField("Count"),0) else numClass = 0 newSinkCount = 0 end if (numClass < 1) then break elseif ((numSinks = numClass) and (sinkCount = newSinkCount)) then break end waterGrid = flowDirGrid.Watershed(sinkGrid) zonalFillGrid = waterGrid.ZonalFill(elevGrid) fillGrid = (elevGrid < (zonalFillGrid.IsNull.Con(0.AsGrid,zonalFillGrid))).Con(zonalFillGrid,elevGrid) elevGrid = fillGrid numSinks = numClass sinkCount = newSinkCount end '*** Rename data set aFN = av.GetProject.GetWorkDir.MakeTmp("ettan", "") '*** Create a theme theGTheme = GTheme.Make(elevGrid) '*** Set name of theme theGTheme.SetName("Filled"++theTheme.GetName) '*** Add theme to the view theView.AddTheme(theGTheme). 40.

(220) theGTheme.SetActive(TRUE) theTheme.SetActive(FALSE) return theGTheme. 41.

(221) ' Name : step2 ' Author : Andreas Gyllenhammar ' Title : Conv_to_grd ' Modified from Spatial.ConvertToGrid and Surface.GridToGrid ' -----------------------------------------------------------------------------' Description: Takes the filled depths GTheme and converts it to a grid ' Runs: As step 2 ' Run by: main ' Self: The current view, the depressionless depth theme. ' Returns: A depressionless depth grid theme ' -----------------------------------------------------------------------------DEBUG = true theView = SELF.Get(0) filleddp = SELF.Get(1) konvlist= List.Make konvlist.Add(filleddp) for each t in konvlist theWorkDir = av.GetProject.GetWorkDir theWorkDir.SetCWD def = theWorkDir.MakeTMP("tvaa", "shp") theDocName = theView.GetClass.GetClassName '*** Set output grid name aFN = "ttvaa".AsFileName '*** Convert selected features of a GTheme to Grid needDelete = FALSE aVTab = t.GetGrid.GetVTab if (aVTab = NIL) then tmpFile = av.GetProject.GetWorkDir.MakeTmp("ctg","") ok = Grid.CopyDataSet(t.GetGrid.GetSrcName.GetFileName,tmpFile) if (ok.Not) then MsgBox.Error("Error during conversion","Convert to Grid :"++t.GetName) return NIL end gridSrc = Grid.MakeSrcName(tmpFile.AsString) aGrid = Grid.Make(gridSrc) needDelete = TRUE else if (aVTab.GetNumSelRecords > 0) then aGrid = t.GetGrid.ExtractSelection else tmpFile = av.GetProject.GetWorkDir.MakeTmp("ctg","") ok = Grid.CopyDataSet(t.GetGrid.GetSrcName.GetFileName, tmpFile) if (ok.Not) then MsgBox.Error("Error during conversion","Convert to Grid :"++t.GetName) return NIL end gridSrc = Grid.MakeSrcName(tmpFile.AsString) aGrid = Grid.Make(gridSrc) needDelete = TRUE end end if (aGrid.HasError) then MsgBox.Error(t.GetName ++ "could not be converted to a grid","Conversion Error") return NIL end status = Grid.GetVerify Grid.SetVerify(#GRID_VERIFY_OFF) if (aGrid.SaveDataSet(aFN).Not) then Grid.SetVerify(status) return NIL end Grid.SetVerify(status) if (needDelete) then Grid.DeleteDataSet(tmpFile) end gthm = GTheme.Make(aGrid) theView.AddTheme(gthm) gthm.SetActive(TRUE) t.SetActive(FALSE) end return gthm. 42.

No results found