1. Steps for the Development of a Model: The case of the Historical Phosphorus Loading Model By Helen Carr

2. Outline  Background  Objectives  Methods  Data  Equations  Conceptual model  Demonstration

3. Description of Problem Lake Champlain has become increasingly eutrophic since the arrival of Europeans 250 years ago. Caused by excess nutrients entering the lake, mainly phosphorus (P) Toxic algae blooms threaten aquatic life and human health Land Use change since settlement is one of the main causes of excess nutrient inputs into the lake.

4. How do we know productivity is increasing? Paleolimnology Colleting sediment from the lake can tell us the trophic history of the lake Sediment cores have been taken from 7 locations in the lake Cores were sectioned, dated and analyzed for P and N accumulation rate and algal biomass.

5. Lake Champlain Basin History 1000 B.C. - 1600 A.D.: Native Americans begin farming Population : 4,000 7000 B.C.- 1000 B.C.: Native Americans hunting and gathering society 1820- 1890: Logging, charcoal and potash production 1870-1900: Population growth slowed forests begin to return 2000: Basin population reaches 571,000, forest covers almost 70% of VT 1870: Maximum deforestation reaches 70% in VT Present 1760- 1800: Period of rapid population growth. 1824-1850: Shift from subsistence to sheep/dairy farming 1609: Samuel de Champlain explores

6. What is causing these problems? Underlying causes of productivity rise in Lake Champlain can be inferred from anecdotal evidence, but quantitative data are lacking

7. Objectives 1) Estimate the total phosphorus loading into Lake Champlain over the past 250 years 2) Quantify the impact of four land uses; cropland, pasture, urban and forested 3) Assess the impact that land use changes such as the period of deforestation and the commercialization of farming have had on the P loading

8. What does it do?? Simulates historical P loading to Lake Champlain based on land use change, atmospheric P deposition, and point sources. –Runs an annual time step from 1760-2010 –Both spatial and temporal resolution are coarse

9. Methods Validate model Develop model Input data Collect data on land use, coefficients, point sources Gather sediment core data Extrapolate and format data Run Simulations Research previous models in literature Collect current P loading values Test model

10. Thinking process  Step 1: to get total Phosphorus loading  Identify all sources of P to lake and create an equation summing all inputs  P load = (Coeff * LU area)+ PSI+ AI+ SI  Input into Simile and test  Step 2: Relate total P load to amount of P deposited  Used total P load and input into a lake compartment  Little research on this so I used a percentage and calibrated to the core data  Step 3: Relate the amount of P to algal growth  Still in progress

11. Data  Model drivers  Land use data 1750-2000 – HYDE database (History database of the global environment)  Land Use data 1992-2001- (Troy et al. 2007)  Point source data- Industrial and sewage- Eric Smeltzer VT DEC  Coefficients Land UseP export*Atmospheric ** Cropland0.250.125 Pasture0.140.125 Urban1.480.26 Forest0.0160.07 *Troy et al. 2007 ** Reckhow et al. 1980

12. Validation Data: Step 1  Phosphorus loading data  Calculated P load from all streams from 1991 to 2008 – Eric Smeltzer VT DEC

13. Validation Data: Step 2  Actual sediment core data from 7 locations  Beta-carotene was used as a indicator of total algae accumulation  Data were determined for each decade and subsequently averaged over the entire lake Phosphorus Accumulation Data Algae Accumulation Data

14. Conceptual Model Phosphorus in Lake Phosphorus in Lake P Loading Atmospheric coefficients Atmospheric coefficients Algal Productivity Algal Productivity Land Use Data TP in sediment core TP in sediment core Outflow Deposition Validate Predicted deposition Predicted deposition P concentration Algae growth Grazing Deposition Validate Algae deposition B-carotene in sediment core B-carotene in sediment core P runoff coeffs Point source data Atmospheric data

15. Equations  Sum ([X]) Sum of numeric array or list  Ex:sum([P_land_use])+sum([P_atmos])+Industrial+Sewage  Min (X,Y) returns the lower value of X or Y  If… Then… Else…  ex: if Add_Intensive==0 then 1 else [Intensive_coeff]  Element ([x],I) Picks the I'th value form the array [x]  Ex: element([10,20,30,40],index(1))  gaussian_var (X,Y) Returns a sample from a Gaussian distribution with mean X and SD Y  Ex: daily_rainfall = gaussian_var(annual_rainfall/365, 1.0)  Use Help > Working with Equations > Built-in Functions

17. Tips  What are your research objectives?  Define them and make sure your model addresses them  Data  What data are readily available?  Units are important!  Make sure they agree  Document everything!  Use the comment and the documentation sections within your model to explain what you did.  Be Organized  Keep your model and your data as organized as possible.

18. Demonstration