Biotechnology.uni-koeln.de

Ecosystem Service and Sustainable Watershed Management in North China International Conference, Beijing, P.R. China, August 23 - 25, 2000 VISIT OF THE REFERENCE AREA IN SANGHUA JIANG
CATCHMENT SOUTH-EAST OF JILIN CITY, NORTH CHINA
- Field Trip -
Armin RIESER (1), Britta STOECKER (2) and Jean-Jacques FORTIER (3)
(1) University of Bonn, Institute of Water Resources and Land Improvement, Faculty of Agriculture, Bonn, Germany, E-Mail ax: +49 228 73 2619 (2) University of Cologne, Institute of Botany, Koeln, Germany Abstract
On August 26-29, 2000, a 4-day field trip followed the Beijing Conference for inspection of the reference area of Songhua Jiang catchment to provide the opportunity to get insight into the local conditions. This was seen as an option for those conference participants who were intending to develop a specific project proposal. The present report of the field trip is based on minutes recorded by Armin Rieser and information added by Mrs Britta Stoecker. Jean-J.
Fortier has added his view how remote sensing and modelling are valuable tools to support RTD projects on environmental and water-related issues. At the end of this report a collection of photographs (Figure Groups 01-10) should provide an impression on Changbai Shan Keywords agriculture, Changbai Shan Mountain, Changbaishan Biosphere Reserve,
ecotourism, Heavens Lake, industry, mining, Sky Lake, tourism impact, vegetation types, Participants:
Prof. Dr. Limin Dai (Shenyang Institute of Applied Ecology, Chinese Academy of Sciences, Prof. Dr. Susete Martins Dias (Centro de Engenharia Biologica e Quimica, Instituto Superior Dr. Jean-Jacques Fortier (SMA Inc., Quebec, Canada)Prof. Dr. Andreas Klik (Institute of Hydraulics and Rural Water Management, University of Dr. Reinhard Perfler (IWGA-SIG, University of Agricultural Sciences, Vienna, Austria)Prof. Dr. Armin Rieser (Institute of Agriculture, Water Engineering and Land Improvement, Britta Stoecker, M. Sc. (Institute of Botany, University of Cologne, Cologne, Germany) In the field staff members of the research station of the Changbai Mountain ForestEcosystems, the Chinese Academy of Sciences, Erdaobaihe, Antu County, accompanied theabove named group: Dr. Gao ChenDr. Hongbing DengBaoying Xiao, M. Sc.
(all members of the Shenyang Institute of Applied Ecology, Chinese Academy of Sciences,Shenyang, PR China) The field trip was guided by Prof. Dr. DAI Limin, the Deputy Director of the Research Stationof the Changbai Mountain Forest Ecosystems, CAS.
Jean-Jacques Fortier - Introductory note - In general, in the following comments, I will try
to evaluate the potential of a GIS system fed by an optimal combination of high and low
resolution satellite remote sensing images and other data. The low resolution satellite (e.g.
Landsat 7) is considerably cheaper, but does not resolve smaller objects (cars, stock-piles,
etc.). To obviate the high costs of the high resolution satellite (e.g. Ikonos), a recourse to
properly designed statistical sample of images (stored as a disconnected GIS layer), can
provide desired quantitative estimates of phenomena, but not a total mapping.
Also, I will stress some possible basic elements of a structure for the GIS, suggested by ourvisit of the reference area. One such important element, it seems to me, is a quantitative waterflow input-output model. The GIS should be structured so as to make possible a calculation,on an on-going basis, of the quantified and qualified input at each user‘s (or group of users)point (e.g a municipality domestic input), as well as the output in the same terms, after use.
As well, each input should be related quantitatively to its source (e.g. a reservoir). In turn asource could be related to another source (e.g. a reservoir to a river, a river to a lake and/or acatchment area). The sources and the users thus form a network which is amenable to aquantitative analysis or a geographical representation.
SATURDAY, August 26th, 2000
Flight with China Northern Airlines, CJ 6152, 11.45 a.m. from Beijing to Yan Ji, transfer withmini-bus to the Changbai Mountain Forest Ecosystem Research Station at Erdaobaihe.
On the way several stops for inspection of the respective catchment area e.g., coal mining,Ginseng plantation. For cultivation of Ginseng the woodland is cleared. The Ginseng can beharvested after 4-5 years. For the reason of reduced nutrients after 5-10 years Ginsengcultivation is followed by crop by use of fertilizer.
Arrival at Erdaobaihe in the late afternoon. - Dinner at the station.
To take a look at some photographs from travel between Yan Ji Airport and the ResearchSt Comments by Jean-Jacques Fortier
I was struck by the seemingly abundant coal mines and coal stock-piles. How many are stillactive, and consequently how much coal is produced and/or stocked domestically andindustrially? If it is possible to locate and delineate stock-piles with low resolution satellite(e.g. Landsat 7), that information could be put in a GIS layer covering the whole case studyarea. Otherwise, a well chosen sample of images with high resolution satellite (e.g. Ikonos)might provide good statistical estimates of quantities of coal stocked on the ground, if not atotal mapping. I assume that the burning of coal has a non negligible adverse effect on theclimate, on the quality of water, on forests (acid rain), etc.
There seems to be several ginseng plantations on the slopes of mountains. The crops takemany years to mature under shading devices that would be recognizable by remote sensing.
Here, a digital elevation model (DEM) incorporated in the GIS, as is customary, would be ofhelp. Aside from providing elevation, it permits the calculation of the approximate slope andaspect of each pixel. These variables help characterize each class of pixels when joined withspectral values to form specific « signatures ». For instance it seems higher sloping ground ischaracteristic of ginseng plantations, very high elevation is characteristic of tundras, etc.
Therefore the presence of these characters facilitate estimation and classification of pixels.
Note : Is there a DEM for the catchment area available at the Institute of Geography? Here again a GIS layer, comprising primary and secondary roads, forest roads, railroads, is ofinterest especially for forest resources accessibility.
Also, a GIS layer for water networks (above and under ground) would obviously be of greatbenefit. Data can be obtained by digitizing maps and plans.
Some other data for a GIS, like industrial sites, small water reservoirs, detailed land use ingeneral, could be obtained from high resolution satellite like Ikonos if only for typical areas.
It is interesting to note that with an Ikonos appropriate sample of images we probably couldmeasure, visually or automatically, and record vehicular activity on the roads (for Dr.
Giancarlo Simeone’s project in particular), activities in forests, etc. (see Ikonos panchromatic1 meter resolution demo image of a section of London below: one can use the « zoom » in thetool bar to view details) Demo I mag e of a S ecti on of L ondon
Changbai Mountain Forest Ecosystem Research Station
In 1979, the Institute of Forestry and Soil Sciences (presently the Institute of AppliedEcology, CAS, Shenyang) was entrusted by the Chinese Academy of Sciences to establish theresearch station for long-term forest ecological research. The Research Station joined theMAB program in the same year and was selected as an "Opened Research Station" by theCAS in 1989. Since 1993, it is instrumental in the International Long-Term EcologicalResearch Network (LTER).
The Research Station is located on a level of 736 m on the northern slope of ChangbaiMountain within the National Reserve and Erdaobaihe Town, Antu County, Jilin Province.
The geographical co-ordinates are 128°28´ E and 42°2´ N.
Research facilities are:• The Station provides laboratories on an area of 700 m² for various analyses.
• The forest hydrological simulation laboratory supports investigations with several • Besides the climate station on a 62.5 m high tower at the Research Station, computerized meteorological stations located in different vegetation zones allow meteorologicalmonitoring.
• Accomodation for guest researchers is offered.
Changbaishan Biosphere Reserve
The Changbai Shan Biosphere Reserve stretches along the boundary between China andKorea between 41°41´49" - 42°24´18" N and 128°16´48" E. It covers an area about 200,000hm². The elevation in that region of volcanic origin ranges from 500 m to the Baiyun peakwith 2,691 m. In the crater of Baiyunfeng the Tianchi (Sky Lake) with an area of 9.4 km² islocated on an elevation of about 2,190 m. The Sky Lake is the main tourist attraction inChangbai Shan Biosphere Reserve. Three access routes go to Sky Lake. The route on the westslope opened in 1995 and the southern route isn´t opened for public up to present (dated in1998). The most frequented route since its opening for tourists in 1985 is the one on thenorthern slope; here the main attractions such as Sky Lake, the route to the Sky Lake, smallSky Lake, hot springs and Changbai waterfall are well developed and managed for touristicobjects.
In 1996 about 226,600 tourists visited Changbai Shan Biosphere Reserve. Because of theclimate (snow cover from October until May) high season is in the summer.
Following environmental impacts have been documented: • garbage pollution: solid waste at touristic attractions (decreasing) • soil erosion: the surface material of Changbai Shan has a loose structure and is prone to • water contamination: is caused by sewage from restaurants and hot spring bath houses.
Along the main tourism routes and scenic sites 4 hot spring bathhouses, 7 toilets and 31garbage boxes are found, which pollute the environment (Source - Yang Xiu & WU Gang (1998) The strategy for conservation and sustainableutilisation of biodiversity in Changbaishan Biosphere Reserve. Journal of Forestry Research 9(3), 217-222; WANG Xianli et al., (1998) Impacts of tourism and developmental potential ofecotourism in and around the Changbaishan Biosphere Reserve. Journal of Forestry Research9 (3), 211-216) on Changbaishan Biosphere Reserve and Changbai Mountain Forest Ecosystem ResearchStation, click at on vegetation types and vegetation zones in Changbaishan Biosphere Reserve, click at on Heavens Lake, waterfall, hot springs, landscape and impact by tourism, click at on research in Changbaishan Biosphere Reserve by Changbai Mountain Forest EcosystemResearch Station, click SUNDAY, August 27th, 2000
At 9 a.m. trip to the upper border of the Songhua Jiang catchment in the forest reserve area, visit of the Sky-Lake respectively the Heavens Lake.
Comments by Jean-Jacques Fortier
There are many different forest compositions sharply separated according to elevation. Itwould be of interest, using remote sensing temporal images like Landsat (that may beavailable from the research center) and the DEM, to measure, with appropriate software likeSMA’s OVERVIEW, the dynamical evolution of the ecological system (on the short term forstress and other natural or man-made events, and on the long term for compositionmodifications of forest cover types).
Of course, to follow the evolution of this large mountain reserve, past and future temporallayers of a GIS system, would have to be used or extended, if already in use.
This data bank on evolution could be augmented and correlated, in the future, with watercatchment quantity and quality measurements obtained by different actions, like monitoringthe Songhua Jiang at its exit from the reserve.
The Sky Lake itself, above the forested area in the tundra, could be (or has been?) the subjectof study, being at the source of the Songhua Jiang, in order to evaluate its effect on waterquantity and quality.
Of course, many prize winning studies have been performed at the center on related subjects.
They would be indicative of how remote sensing and GIS could be configured. Also, theywould tell which ground truth control points of forestry and other cover classes should beidentified. These would be georeferenced by ground positioning systems (GPS) if identifiedon the terrain, or directly, if identified on an aerial photograph or on the satellite image.
2. Visit of the meteorological observation point, altitude 2,300 m (lowest temperature duringthe last night: 2 degree Celsius) and inspection of the nearby permanent plot No.4 in themountain tundra. The permanent plots are located in different vegetation zones and have asize of 1 ha; investigations are on species and density of vegetation, soil, litter and meteorolo-gical conditions.
Comments by Jean-Jacques Fortier
I have had experience with SMA’s software SURFACES in the analysis of vegetation throughcolor numerical images, taken with numerical cameras held approximately 2 meters aboveground. The advantage of the method is that it is non destructive of the vegetation or the litter,so one can follow its evolution through a season or more. It is also time and money saving.
Classes of vegetation can be measured semi-automatically by the computer, once the sampleof images is acquired, and classes located and identified. Possibly, litter can also be classifiedand measured.
If other measurements, like temperature, soil humidity, depth of litter, are simultaneouslyrecorded with the images, correlations can be established with the vegetation and littercontents.
Another application would be to take samples of images in exact correspondance to theground control points for tree classes in forested areas. These would be used to supplementsatellite images that cannot see the ground vegetation and litter under the trees.
Meeting with Liu Xiang, Department of Geology, Jilin University, Changchun and Alessandra Esposito, Guido Giordano, Donatella de Rita, Dipartimento Scienze Geologiche,Universita Roma Tre; discussion on the geology of the upper catchment (last eruptions of thevolcanoe 1597, 1688 and 1702).
5. Visit of the 68 m high waterfall, drainage of the Sky-Lake to the Songhua, other drainageis to the area of North-Korea.
Comments by Jean-Jacques Fortier
See the comments above regarding the Sky Lake (in the French guide this lake is called « Lacdu Paradis »).
Walk to the forest sampling site near Sanheshui Power Station, altitude 850 m.
The sampling sites by Dr. Hongbing Deng, working on the nutrient cycling in terrestrial andaquatic ecosystems, are located in an interval of 50 m altitude (altogether 20) and have a sizeof 100 m in the length and 8 m in the width; investigation are on vegetation and soil water;amount and contents of run-off and stem-flow; litter at 6 samplings (each 1 m²) in an area of900 m²; water and forest productivity.
Comments by Jean-Jacques Fortier
Possibly, the same comments apply as in permanent plot No.4, regarding the use of images
taken with a numerical camera. Though only canopies are recorded by the camera, a larger
number of samples of 1 m2 as in this experiment, can be taken, due to the rapidity and simpli-
city of photographing chosen plots.
The number and location of image sites could also be determined from a statistical analysis ofthe variance of the contents of a first sample of the images. This statistical approach with thecamera could be applied to the present method of laying traps in the forest, if camera investi-gation is not sufficient.
7. Inspection of the Hydrological Simulation Laboratory of the Research Station and anexperiment by Mr. Wu, M.Sc.-Student from Prof. Pei, Institute of Applied Ecology,Shenyang-University on soil erosion and subsurface flow in 8 depths (altogether 1,2 m deep).
Size of the test: 2,8 m x 5 m, variables: rainfall depth and intensity, soil, slope and treespecies.
Comments by Jean-Jacques Fortier
It was said that such a large apparatus is a very ambitious one not available elsewhere. Icannot see, at this stage, how GIS or remote sensing, or other methods I have mentionedcould be useful to this experiment For view on Hydrological Simulation Laboratory MONDAY, August 28th, 2000
At 9 a.m. inspection of the outlet of the Erdaobaihe-town sewage system (water is drained without purification into the Songhua).
Comments by Jean-Jacques Fortier
Obviously one could add to the GIS layer of water networks, like mentioned in 1. above, alayer of sewage networks, at least, in both cases, for the main conduits. Characteristics, likecapacity, type of treatment, are natural add-ons in the network layers.
2. Visit of Shui Yun Di water supply of the Erdoabaihe municipal water supply authority bynatural springs in a secondary forest area. The inner water protection zone is well fenced. The total water flow is about 10,000 m³ for the 30,000 inhabitants of Erdaobaihe. The water is nottreated.
Comments by Jean-Jacques Fortier
This natural spring is probably detectable from satellites, at least from high resolution ones.
An interesting problem is to determine how many km2 of seemingly humid forest arenecessary to catch enough water to produce, in a spring, the present capacity (10 000 m3).
To start answering the question, one might determine from satellite observations how manysuch natural springs (or water surfaces) there are per square kilometer, for forested, on the onehand, and unforested area, on the other hand. Possibly, a quantitative analysis from a GISlayer of such springs or surfaces, will indicate where and how much of this natural springwater is available. (This permits correlation between natural spring water and the type ofterrain, and importance of forest cover) To take a look at some photographs concerning to Erdaobaihe drinking water provision andwaste/sewage management, click at Walk to a large open wetland area, east of Erdaobaihe.
Comments by Jean-Jacques Fortier
This vast wetland would be easily identified with low resolution satellite and would beentered in another important layer of the GIS. Also we could identified peatlands fromsatellite images like we did in Quebec, with SMA‘s software module PEATLANDS, and formanother GIS layer.
Again, applying parameters of water catchable per type and per m2 of wetlands, one couldcalculate from the GIS layer the global capacity of wetlands in each area to deliver unpollutedwater.
Inspection of an agricultural area about 80 ha for 40 farm families, east of Erdaobaihe, newly cultivated with corn, soya, sorghum, sunflower, potatoes. In the housegardensadditionally cucumber, eggplants, beans, pepper, mushrooms, tobacco. The farmers keep asmall number of chicken, cows and pigs.
In one of the farmyards a shallow groundwater well with a pump could be seen for thedrinking water supply.
Comments by Jean-Jacques Fortier
In Erdaobaihe, as elsewhere, there are many discrete small fields of corn, and other crops
(sunflowers, canola, etc). One problem for measurements of such fields (e.g. surfaces of each
crop), with economical low resolution satellites like Landsat, is the large size of the pixels
(900 m2) of such low resolution satellites.
This creates so-called « mixed pixels » that overlap different fields at their frequent sharpborders. And, therefore, smaller fields tend to be underestimated, or even erased from the In view of this, I would resort to a technique, MIXPIX, I developed for mixed pixels, intreating the farm layer of the GIS. This layer, on the other hand, would be obtained byisolating the farmlands from the rest of the landscape, by a technique used in thePEATLANDS software mentioned above.
One could then calculate what sort of surface or ground water consumption and/or pollution isentailed by each crop per unit of acreage.
For more information on agriculture, click at Aquaculture (200 x 200 m, 6 basins) for fish-production.
Inspection of wild and widespread garbage-pollution in the wetlands where sand was Comments by Jean-Jacques Fortier
It is possible to distinguish the remotely sensed spectra of those garbage polluted wetlands asopposed to virgin wetlands and, then, to estimate the number and areas of such pollutedwetlands as a proportion of total. Of course it would be possible to follow the evolution of thephenomenon through a temporal succession of images (past or future).
Note: There is a problem with temporal analysis using a succession of past images. Theseimages not being spectrally equivalent,due to the different atmospheric conditions, they mustbe recalibrated with contemporaneous control points, often not available. So, from a presentsample of control points,one must evaluate (by considering features of those images at thesame points) whether the control points still contain the same classes (with possibly an agefactor subtracted: e.g. the control point is estimated to contain the same class of trees as now,but possibly younger ones if the point is known to be in regeneration). 7. Visit of Baishan Shengsui table water factory (15.000 bottles per day, 600 ml each);purification of the municipal drinking water by green sand and activated carbon.
Comments by Jean-Jacques Fortier
Here, as elsewhere, the natural question is about the statistics of such consumption. Also aGIS mapping of such consumption by locality (e.g. using sales documents as base data) wouldbe of interest. One could compare with tap water drinking consumption, which would bemeasured as a fraction of the water network input.
Inspection of a pressed wood factory sewage water outlet, going without purification into Comments by Jean-Jacques Fortier
Again, one is in need of a quantitative analysis. Data on water input and sewage outputquantity and quality, for such a factory, other factories, municipalities, farm areas, etc., wouldpermit a global analysis of water input and output, using the GIS and the water flow modelmentioned before.
9. Visit of a Sawmill of the Bai He Forestry Bureau. The outer part of Changbai ShanBiosphere Reserve is an important timber production base in China.
Comments by Jean-Jacques Fortier
The quality of the wood (non uniform shape, damage, rot) precludes full computer control ofthe saws, like Quebec forest companies are familiar with. Is such wood doubtful quality dueto the preservation of good quality forests, or is it the only available wood in the area? In general, the exploitation of forests for supply to the sawmills, paper mills, pressed woodfactories, would be analysed using a temporal succession of GIS layers, provided by remotesensing of past cuts in the forests, and their measurements by classes.
What is the correlation between forest exploitation with water quantity and quality. GIS calcu-lations might provide some answers. This would require the establishment of a dynamicalmodel relating all the elements mentioned before, in view of creating a kind of water input-output table « à la Leontieff ».
10. Guided visits of Fu Bai Wood Company, pressed wood factory, state enterprise of theBaihe Forestry Bureau.
11. Inspection of an outlet of the hospital of Erdaobaihe. The sewage water from the hospitalis drained directly only after a mechanical purification into the river.
Comments by Jean-Jacques Fortier
See above for sewage treatment. Such constructions (schools, office buildings, etc) whenregistered in a GIS with occupancy data and other data, would provide quantitative elementsfor the water input-output suggested model.
12. Visit of the Natural Sciences Museum of Changbai Mountains in Erdaobaihe.
After that manifold program and a small supper the task-force for the field visit was invited by Prof. Dr. Limin DAI to a Korean barbecue during that all the participants expressed theirsincere thanks.
The finale was an extended Karaoke in the meeting hall of the research station until 11 p.m.;for most of the Non-Chinese participants the first time in their lifes.
For some more photographs on landscape and community, click social events, click at TUESDAY, August 29th, 2000
Back to Yan Ji to the airport with mini-bus. Several stops on the way for inspection ofinteresting sites. Passing of the Antu drinking water reservoir (no protection measures visiblein the protection zones) and stop at Antu, where is the China Water Authority for the wholeAntu-County.
Comments by Jean-Jacques Fortier
It seems it is used for drinking water (and other domestic and industrial necessities?). Howmuch water is there for domestic or industrial consumption? In general how much water is stored in reservoirs for different uses like domestic con-sumption, electricity generation, irrigation, etc. It would be possible to register in a GISsystem (as a time series?) water reserves by areas, with the help of remote sensing andexisting maps, and other known georeferenced data.
FINAL NOTES :
The visit has allowed us to examine most of the catchment area‘s elements of interest: roads,farms, villages, forests, tundras, rivers, lakes, springs, industries, hospitals, waterworks,sewage and refuse outlets, wetlands, scientific measurement endeavours, etc.
To construct a GIS, one needs, in my view, a water utilization model incorporating qualifiedand quantified water input and output of different entities: municipalities, industries, hospi-tals, office buildings, farm areas, etc.
This model, in turn, would serve the watershed management interventions (e.g buildingspecific constructed wetlands, having verified experimentally their capabilities, and knowing,from the model, the effects of the industrial use on water quality and quantity).
If high resolution satellites (e.g. IKONOS) global coverage, as opposed to sampling, is neces-sary, the GIS could not cover an extensive area, unless ample financial resources are availableto the GIS project.
In general, from the information needs in each project, we could determine the coverage ofhigh resolution images necessary, and, consequently, establish some preliminary limits to thecase study GIS catchment area.
Also, an investigation of the availability of reasonably up-dated georeferenced data, locally orin central government offices, in digitized or digitizable form (like maps), should precede theestablishment of the GIS, and set the number of the water model variables. (Digitizing can be All in all, it seems that we should choose a small enough but representative catchmentreference area, and a parsimonius choice of water model variables, in view of the cost ofobtaining the information. In this respect, the ChangBai forest reserve and the ErdaoBaiheregion seem ideal by their diversity, though there is no large city in the area.
Flight back to Beijing with China Northern Airlines, instead of CJ 6151 with CJ 6163,because of delay, 6.50 p.m., arrival at Beijing-airport 10 p.m., and with taxi to the XiyuanHotel, 1 Sanlihe Road, Beijing.

Source: http://www.biotechnology.uni-koeln.de/inco2-dev/common/contribs/44_field.pdf