Overview The overall goal of our research project

Overview: • The overall goal of our research project is to develop an integrated remote sensing simulation system that more accurately accounts for CO 2 fluxes between agricultural lands and the atmosphere. • Goal 1 of our project is to develop a new decision-support assessment tool incorporating remote sensing imagery into the inventory analysis for improving crop and grazing land NPP estimations. • One preliminary step is to carry out model-data net primary production (NPP) intercomparisons for a variety of agricultural systems – various crops, locations, and management. • Here we present: • information about the Century and CASA (Carnegie-Ames-Stanford-Approach) models and survey data from the National Agricultural Statistics Service (NASS); • the data we use to drive them; and • comparisons of annual and monthly NPP data from the three sources. Methods • • • Sub-county modeling in both Iowa and Nebraska Results summed and compared at the county scale (each county is 1 point) Time period: 2001 -2004 Primary crops: corn, soybean, wheat Irrigated versus non-irrigated corn and soybeans Compared with data derived from National Agricultural Statistics Survey (NASS) yield reports for the same study period (ask about details). The challenge: Our goal is to develop a new model that integrates a model with observation-observed NPP constraints (CASA) with the most robust description of soil organic matter dynamics (Century). We anticipate that the new model will allow us to more accurately and confidently The main challenge to integrating these two models is that the Century soil carbon contents and the distribution across carbon pools are a function of plant productivity, plant water uptake, and plant constraints on nutrient cycling. Substitution of external rates of NPP – like those derived from CASA – could lead to imbalances in potential evapotranspiration (PET) and plant-available soil moisture and nitrogen demand uptake. Since most Century parameterizations tune model output to soil C stocks, substituting external rates of NPP could also lead to mis-estimation of soil C stocks. The Century model NASA-CASA model The Century ecosystem model simulates plant and soil C, N, P, and S dynamics. Flows of C and nutrients are controlled by the amount of C in the various pools (e. g. , soil organic matter, plant biomass), nutrient concentrations of the pools, abiotic temperature/soil water factors, and soil physical properties related to texture. Century runs on a monthly time step. NASA-CASA is an ecosystem model that simulates carbon and nitrogen transformations. Net primary production (NPP) is estimated in the model by photosynthetic efficiency functions. The model includes carbon flux controls of nutrient substrate availability, soil moisture, temperature, texture and microbial activity. Monthly NPP is estimated as a product of time-varying surface solar irradiance, Sr, and enhanced vegetation index (EVI) from satellite remote sensing, plus a constant light utilization efficiency (LUE) term (emax) that is modified by time-varying environmental stress terms for temperature (T) and moisture (W) effects. Lignin content • There are several counties for which data sources disagree about whether crops are grown or not. • CASA and Century soybean and irrigated corn estimates agree better with NASS NPP estimates than estimates for other crops. • CASA and Century both tend to overestimate NPP for counties that NASS estimates low NPP. • CASA NPP estimates are generally greater than Century estimates for all crops. CO 2 Structural C The emax LUE term was set uniformly at 0. 55 g C MJ-1 PAR for C 3 plants and 1. 66 g C MJ-1 PAR for C 4 plants, values that were derived from calibration of predicted annual NPP to thousands of previous NPP field measurements (including the Agro 2002 comparison carried out at the corn/soybean site Bondville, IL). Monthly, 8 -km gridded satellite greenness products were from MODIS (Moderate Resolution Imaging Spectroradiometer) between 2001 and 2004. Monthly mean TEMP and PREC grids for model simulations over the years 2001 -2004 came from NCEP reanalysis products. Active C CO 2 Slow C CO 2 Passive C CO 2 Wetter dryland crops Corn/soybean Annual NPP comparisons: NPP = Sr * EVI * emax * T * W. metabolic C Drier Irrigated+dryland Wheat Corn Soybean Wheat NASA-CASA Model Design The grassland/crop production model simulates plant production for different herbaceous crops and plant communities (e. g. warm or cool season grasslands, wheat and corn). The plant production model assumes that the monthly maximum plant production is controlled by moisture and temperature and that maximum plant production rates are decreased if there are insufficient nutrient supplies (the most limiting nutrient constrains production). The fraction of the mineralized pools that are available for plant growth is a function of the root biomass with the fraction of nutrients available for uptake increasing exponentially as live root biomass increases from 20 to 300 gm-2. Most forest or grassland/crop systems are limited by nutrient availability and generally respond to the addition of N and P. Plant growth is affected by the structure of the soil as specified in the site parameters. The model runs for this exercise were based on the same driving data used for US National agricultural soil C inventory, namely USDA-Natural Resources Inventory Land use data, PRISM gridded climate data, and STATSGO/NRI soil data. Regional land use histories were used for model spin-up. (a) Soil Moisture Balance PPT f(TEMP) f(WFPS) f(SOLAR) PET Soil Surface M 0 M 1 Heat & Water Flux (b) Ecosystem Production Nutrient Mineralization M 2 M 3 Grass/Crop FPAR CO 2 NEP NPP Biomass Soil Leaf Litter Profile Root Litter Layers Microbes f(TEMP) f(WFPS) f(Lit q) Soil Organic Matter Rh CO 2 f(Lit q) Monthly NPP comparisons: • CASA monthly NPP estimates tended to be greater than Century NPP estimates, for all crops: dryland (average 100%) and irrigated corn (36%), soybeans (42 and 25% for irrigated and dryland), and wheat (76%). • Estimates from the two models differed by the largest proportion during the early (MAM) and latter (SON) parts of the growing season, though differences during the summer months (JJA) contributed the most to differences in NPP: 47% of total difference for corn, 84% for soybeans, and 58% for wheat. The approach: • We plan to develop a model that shares a water balance model similar to the simple model used in Century, but that accurately replicates the water submodel within the CASA model. • We will approximate the plant N uptake using an optimization routine in which plants (based on NPP from NASA-CASA) remove N from the soil mineral N pool after microbial immobilization, with the potential for a feedback if there is not enough nitrogen to meet the stoichiometric requirements of the plants (i. e. , N immobilized will be re-allocated to the plants). • The NPP will be prescribed on a monthly basis in Century from the NASA-CASA simulations. • This poster presents comparisons of CASA and Century NPP as a preliminary evaluation of the capacity to prescribe NPP within the linked model using CASA. Next steps: • • • Re-run CASA analyses using 250 m MODIS EVI data, common weather data, and (possibly) SSURGO soils. Carry out similar comparisons to assess agreement between fine-scale CASA output and Century output. Explore how the two models apply limitations due to water and nitrogen and assess temporal and spatial alignment of water and N limitation on NPP To the extent possible, identify areas where corn and soybean area grown and carry out model runs for those areas Conduct site-level model assessments at 28 sites we have identified for which yield and soil C observations have been collected and that are large enough to contain several MODIS pixels.