Systems and Modeling Introduction

Transcription

1 Systems and Mdeling Intrductin What is a system? System: A cllectin f intercnnected parts that functin as a cmplex whle, thrugh which matter cycles and energy flws. System Characteristics Systems are made up f systems The whle is greater than the sum f its parts A system will change if ne r mre parts are remved. A system has qualities that culd nt be achieved simply by adding up its individual parts. Systems can be clsed r pen Systems have feedbacks Examples Human Bdy Cmputer Water Cycle Baseball Team Car Earth Nn-examples Desk Cking pt Rcks Bwl f fruit Bttle f water Bk Systems Vcabulary - Open System: a system that cntinuusly interacts with its envirnment Transfer f infrmatin, matter, r energy acrss the system bundary Implies there are inexhaustible amunts f energy available t the system - Clsed System: a system that is self- cntained in regard t transfer f matter N matter crsses the system bundary, but energy des We have t use this definitin because there is nly ne truly clsed system the Universe (t ur knwledge) - Feedback: A respnse t sme actin within the system. The linkage between tw r mre system parts that frms a rund- trip flw f infrmatin is thus called a feedback lp. (One event causes anther, and the secnd event cmes back arund t influence the first.) Negative Feedback (balancing feedback): When an actin in ne directin eventually causes a balancing reactin in the ppsite directin. Examples t remember include the heating system in yur huse r maintenance f yur internal bdy temperature. (It has nthing t d with gd r bad.) Psitive Feedback (reinfrcing feedback): A respnse that keeps the system mving in the same directin, amplifying the effect. This type f feedback is cmmnly assciated with expnential grwth, such as in a ppulatin. 1 f 5 Systems & Mdeling Intrductin 2011 A cllabrative prject between the University f New Hampshire, Charles University and the GLOBE Prgram Office.

2 - Pl (als stck r reservir): A pl is the strehuse f material in a prtin f the envirnment. Examples f 'pls' scientists might cnsider include: carbn in leaves, trees r entire ecsystems; water in a river, lake r all f the wrld's ceans; calcium in rcks, seashells r yur wn bdy. Scientists use the cncept f a pl as a way f simplifying what wuld therwise be very difficult t study. - Flux (als flw): The mvement f material frm ne pl t anther, per unit time. (Fluxes are ften a prcess, such as phtsynthesis r evapratin.) Input: The flw f material entering a pl Output: The flw f material leaving a pl - Steady state (als dynamic equilibrium): A cnditin in which the amunt f material within the pl at any time remains the same. In ther wrds, the rate f input equals the rate f utput, causing the ttal pl size t remain unchanging in time. - Turnver rate: The fractin f material that leaves a pl in a specified time interval. Turnver rate is the mathematical inverse f residence time. - Residence time: The average length f time that material spends in a given pl. Residence time depends n the rate f utflw and n the size f the pl. Residence time is the mathematical inverse f turnver rate. Systems Diagrams It is ften easiest t understand a system thrugh visual representatin. Here is a 1- bx stck and flw diagram t shw the basic system cmpnents. Once a basic system has been represented mdelers add feedbacks, and define relatinships using equatins s the system can be bserved ver time and under varying cnditins. Input Flw, Flux Per unit time Pl, Stck, Reservir Output Flw, Flux Per unit time What are mdels? When asked this same questin, a grup f GLOBE Carbn Cycle teachers defined mdels in their wn wrds: A mdel is a tl used t refine ur understanding f the wrld, ask new questins, and make infrmed predictins. 2 f 5 Systems & Mdeling Intrductin 2011 A cllabrative prject between the University f New Hampshire, Charles University and the GLOBE Prgram Office.

3 A mdel is a small r large- scale representatin f a prcess that ccurs n earth that might be hard fr us t study n its natural smaller r larger scale. A mdel is a replica f an bject r cncept that helps increase ur understanding. A mdel is a representatin f reality that can take many frms, physical, cnceptual and/r mathematical. It has sme characteristics f reality, but has ther characteristics that wn t match reality. Mdels are ften used fr testing purpses t evaluate different scenaris. Fr the GLOBE Carbn Cycle Prgram we define mdels as: tls that help us understand, explain, and predict systems that are t cmplex r difficult t bserve r cmprehend n ur wn. We ften think that mdels must include cmplicated mathematical functins, but we can learn a great deal abut systems, especially envirnmental systems, using simple mdels that require nly the basic math functins such as additin, subtractin, multiplicatin and divisin. Why use mdels? Mdels are particularly useful in the study f envirnmental systems. Scientists use mdels t examine the fundamental behavir f a system, such as hw carbn is stred in a frest. By knwing mre abut the system thrugh mdeling and understanding where knwledge is incmplete, scientists can generate hyptheses t guide future research. Mdels are als useful t predict future cnditins. Althugh such predictins are nt necessarily what will happen, they can prvide a range f pssibilities that help scientists refine their research and help plicy makers create actin plans t prevent any undesirable utcmes. As an example we might use a mdel in asking this questin, hw might carbn uptake and strage in frest ecsystems change if there is an increase in temperature ver the next twenty years? Hw are mdels develped? The develpment f mdels is a cyclical prcess. Scientists begin with a questin. They then cllect field data r run experiments t understand relatinships within a system that will help them answer the questin. The infrmatin is then used t build simple cmputer mdels. Scientists cntinue t ask and answer questins by making mdel runs and examining the utcmes. Often mdel utcmes lead t new questins that require mre cmplicated mdels. In rder t answer these questins scientists must have mre knwledge abut the underlying prcesses f the system, which can nly be understd thrugh mre data cllectin (fieldwrk/experiments). As mre and mre detailed relatinships and system prcesses are added t the mdel, the uncertainty we have in the mdel results decreases. Fr example, the carbn strage in abvegrund tree parts (leaves, branches, stems) has been successfully mdeled based n millins f measurements. Measurements allwed scientists t build a mdel that reprduces what the envirnment actually des in ne place at ne particular time. This mdel can nw be used t make best guesses abut hw carbn strage in these tree parts might change if there were changes in light, temperature r precipitatin. On the ther hand, while these mdels were being develped and used it became clear that there wasn t enugh detailed infrmatin available abut the belwgrund parts (carse and fine rts) fr scientists t fully understand hw carbn strage belw the grund changes ver time. In this case scientists have t g back t the field and make mre 3 f 5 Systems & Mdeling Intrductin 2011 A cllabrative prject between the University f New Hampshire, Charles University and the GLOBE Prgram Office.

4 measurements, lking fr clearer relatinships between belwgrund parts and the abvegrund parts and prcesses they already understand. Mdeling Systems - Examples As a simple example, cnsider a High Schl as a system. We will fcus n the mvement f peple (matter) thrugh the system. Cnsider a high schl that has 800 students (pl) and these students are evenly divided amng the 4 grade levels (Freshman, Sphmre, Junir, Senir). If we assume that all senirs graduate every spring, 200 students are leaving the schl (utput flux). Let us als assume 200 new freshman enter each fall (input flux). (Thus ur system is in equilibrium.) If turnver rate is the fractin f students that leave the schl each year (utput flux/pl size), then ur equatin is: (200 students/year)/800 students =.25 This means that 25% f the student bdy is graduating and leaving the schl each year. If residence time is the number f years that a student spends at the schl befre they graduate (pl size/utput flux), then ur equatin is: 800 students/(200 students/year) = 4 years This means that every student spends 4 years in the schl befre they graduate. The methd we used t calculate turnver rate r residence time in the abve example is the typical methd, which can be used fr systems that are either in steady state OR nn-steady state. In sme cases hwever the utputs f a system are hard t calculate at a particular scale, s we must assume the system is at steady state. In this case we can then use the knwn inputs because they are theretically equal t the utputs. This will prvide us with at least an estimate f turnver rate and residence time fr the system. While this methd is less desirable because inputs d NOT actually determine the turnver rate r residence time, in sme cases it is necessary. One example f this situatin is the GLOBE Carbn Cycle Glbal Bimass Accumulatin Mdel. NPP (ttal phtsynthesis by vegetatin minus plant respiratin - determined by temperature r precipitatin) Bimass Litterfall (all f the dead material in a given year leaves, branches & stems determined by turnver rate) Because litterfall is difficult t quantify at the bime- level spatial scale, this mdel uses a turnver rate that is determined by dividing NPP f the bime (vegetatin) by the ttal bime bimass (vegetatin), 4 f 5 Systems & Mdeling Intrductin 2011 A cllabrative prject between the University f New Hampshire, Charles University and the GLOBE Prgram Office.

5 thus inputs were divided by pl size instead f the utputs. T d this we had t assume that the system was in steady state. T get a mre accurate turnver rate fr a particular lcatin within the bime, we wuld have t make measurements f the ttal litterfall during a ne- year perid. This might include putting ut baskets t catch leaves, twigs, and fruit, laying ut tarps t cllect falling branches, and setting up several transect lines t measure trees fallen acrss them. T help students experience hw mdeling (diagramming, equatins) can be used t understand systems have them participate in the GLOBE Carbn Cycle: Paper Clip Simulatin Activity and Cmputer Mdel. References/Resurces Bth- Sweeney, Linda. (2001). When a Butterfly Sneezes. Waltham: Pegasus Cmmunicatins, Inc. Bth- Sweeney, Linda, and Dennis Meadws. (2010). The Systems Thinking Playbk: Exercises t Stretch and Build Learning and Systems Thinking Capabilities. White River Junctin: Chelsea Green Publishing. isee Systems: The Wrld Leader in Systems Thinking Sftware. (2011). Kump, Lee R., James F. Kasting, and Rbert G. Crane. (2004). The Earth System: 2 nd Editin. Upper Saddle River: Pearsn Educatin. Meadws, Dnella H. (2008). Thinking in Systems: a primer. White River Junctin: Chelsea Green Publishing. Quaden, Rb, Alan Tictsky, and Debra Lyneis. (2004). The Shape f Change. Actn: Creative Learning Exchange. Quaden, Rb, Alan Tictsky, and Debra Lyneis. (2007). The Shape f Change: Stcks and Flws. Actn: Creative Learning Exchange. Sussman, Art. (2000). Dr. Art s Guide t Planet Earth. San Francisc: WestEd. 5 f 5 Systems & Mdeling Intrductin 2011 A cllabrative prject between the University f New Hampshire, Charles University and the GLOBE Prgram Office.