Ankur Desai - Academia.edu (original) (raw)
Papers by Ankur Desai
Evapotranspiration (ET) prediction and forecasting play a vital role in improving water use in ag... more Evapotranspiration (ET) prediction and forecasting play a vital role in improving water use in agriculturally intensive areas. Metrological and biophysical predictors that drive ET in managed landscapes have complex nonlinear relationships. Deep learning and data-driven methods have shown promising performance for identifying the dependencies among variables. Here, we evaluated the potentials of random forest (RF) and long short-term memory (LSTM) neural networks to estimate and forecast daily ET for corn, soybeans, and potatoes in diverse agricultural farms during 2003-2019. The modeling framework was applied for nineteen fields where eddy covariance ET and meteorological observations in the Midwest USA for growing season (April-October) is available. In this study, we applied data-driven models (RF and LSTM) with 3 sets of predictors (5, 11, and 16 predictors). Results show that a 16 predictor RF model (RF_16 R 2 = 0.7, Willmott's skill score = 0.90) outperformed a process-bas...
The potential of a stepwise fusion of proximally sensed portable X-ray fluorescence (pXRF) spectr... more The potential of a stepwise fusion of proximally sensed portable X-ray fluorescence (pXRF) spectra and electromagnetic induction (EMI) with remote Sentinel-2 bands and a digital elevation model (DEM) was investigated for predicting soil physicochemical properties in pedons and across a heterogeneous 80-ha crop field in Wisconsin, USA. We found that pXRF spectra with partial least squares regression (PLSR) models can predict sand, total nitrogen (TN), organic carbon (OC), silt contents, and clay with validation R 2 of 0.81, 0.74, 0.73, 0.68, and 0.64 at the pedon scale but performed less well for soil pH (R 2 = 0.51). A combination of EMI, Sentinel-2, and DEM data showed promise in mapping sand, silt contents, and TN at two depths and Ap horizon thickness and soil depth across the field. A clustering analysis using combinations of mapped soil properties or proximal and remote sensing data suggested that data fusion improved the characterization of field-scale variability of soil properties. The cost-benefit analysis showed that the most accurate management zones (MZs) for topsoil can be generated only using estimated soil property maps while it was the most costly as compared to other data sources. For an intermediate-high (for topsoil) and high (subsoil) accuracy and a moderate economic budget, the combination of sensors (proximal + remote sensing + DEM) might be a better approach for effective MZs generation than collecting soil samples for laboratory analysis while the latter produced the most accurate maps for topsoil. It can be concluded that pXRF spectra can be useful for predicting key soil properties (e.g., sand, TN, OC, silt, clay) at different soil depths, and a combination of proximal and remote sensing provides an effective way to delineate soil MZs that are useful for decision-making.
Evapotranspiration (ET) prediction and forecasting play a vital role in improving water use in ag... more Evapotranspiration (ET) prediction and forecasting play a vital role in improving water use in agriculturally intensive areas. Metrological and biophysical predictors that drive ET in managed landscapes have complex nonlinear relationships. Deep learning and data-driven methods have shown promising performance for identifying the dependencies among variables. Here, we evaluated the potentials of random forest (RF) and long short-term memory (LSTM) neural networks to estimate and forecast daily ET for corn, soybeans, and potatoes in diverse agricultural farms during 2003-2019. The modeling framework was applied for nineteen fields where eddy covariance ET and meteorological observations in the Midwest USA for growing season (April-October) is available. In this study, we applied data-driven models (RF and LSTM) with 3 sets of predictors (5, 11, and 16 predictors). Results show that a 16 predictor RF model (RF_16 R 2 = 0.7, Willmott's skill score = 0.90) outperformed a process-based land surface model (LSM R 2 = 0.57, Willmott's skill score = 0.86) for predicting daily ET, while LSTM performance was lower (LSTM_16 R 2 = 0.65, Willmott's skill score = 0.89 and LSTM_11 R 2 = 0.62, Willmott's skill score = 0.86) than RF using the same sets of predictors. Vapor pressure and crop coefficients were identified as the most important predictors for irrigated crops, while short wave radiation and enhanced vegetation index were key predictors for non-irrigated crops. For certain crop types, such as corn and soybeans on fine-grained soils (silt loam), a simpler version RF, using only 11 drivers, can provide comparable results (R 2 = 0.70 vs 0.69 and Willmott's skill score = 0.90 vs 0.88). For short-term 3-day ET forecasting, LSTM is more sensitive to uncertainty in ensemble forecast meteorology than RF. ET forecasts were strongly sensitive to forecast uncertainty of vapor pressure. The proposed modeling architecture provides a field-scale, locally calibrated tool for accurate prediction and short-term forecasting of daily ET in areas where in situ ET, metrological, and biophysical data are lacking.
Sustainability, 2020
Dairy farms are predominantly carbon sources, due to high livestock emissions from enteric fermen... more Dairy farms are predominantly carbon sources, due to high livestock emissions from enteric fermentation and manure. Integrated crop-livestock systems (ICLSs) have the potential to offset these greenhouse gas (GHG) emissions, as recycling products within the farm boundaries is prioritized. Here, we quantify seasonal and annual greenhouse gas budgets of an ICLS dairy farm in Wisconsin USA using satellite remote sensing to estimate vegetation net primary productivity (NPP) and Intergovernmental Panel on Climate Change (IPCC) guidelines to calculate farm emissions. Remotely sensed annual vegetation NPP correlated well with farm harvest NPP (R 2 = 0.9). As a whole, the farm was a large carbon sink, owing to natural vegetation carbon sinks and harvest products staying within the farm boundaries. Dairy cows accounted for 80% of all emissions as their feed intake dominated farm feed supply. Manure emissions (15%) were low because manure spreading was frequent throughout the year. In combination with soil conservation practices, ICLS farming provides a sustainable means of producing nutritionally valuable food while contributing to sequestration of atmospheric CO2. Here, we introduce a simple and cost-efficient way to quantify whole-farm GHG budgets, which can be used by farmers to understand their carbon footprint, and therefore may encourage management strategies to improve agricultural sustainability.
Agricultural and Forest Meteorology, 2007
We review 15 techniques for estimating missing values of net ecosystem CO 2 exchange (NEE) in edd... more We review 15 techniques for estimating missing values of net ecosystem CO 2 exchange (NEE) in eddy covariance time series and evaluate their performance for different artificial gap scenarios based on a set of 10 benchmark datasets from six forested sites in Europe.
Stand-replacing disturbances including harvests have substantial impacts on forest carbon (C) flu... more Stand-replacing disturbances including harvests have substantial impacts on forest carbon (C) fluxes and stocks. The quantification and simulation of these effects is essential for better understanding forest C dynamics and informing forest management in the context of global change. We evaluated the process-based forest ecosystem 5 model, PnET-CN, for how well and by what mechanisms changes of ecosystem C fluxes, aboveground C stocks (AGC), and leaf area index (LAI) arise after clearcuts. We compared the effects of stand-replacing harvesting on C fluxes and stocks using two chronosequences of eddy covariance flux sites for deciduous broadleaf forests (DBF) and evergreen needleleaf forests (ENF) in the Upper Midwest region of northern 10 20 covery. Simulated NEP for both forest types was initially negative with the net C losses of ∼ 400-700 g C m −2 yr −1 for 6-17 years after harvesting, reached the peak values of ∼ 400-600 g C m −2 yr −1 at 14-29 years of age, and became stable and a weak C sink (∼ 100-200 g C m −2 yr −1 ) in mature forests (> 60 years old). The decline of NEP with age was caused by the relative flatting of GPP and gradual increasing of ER. ENF re-25 covered slower from net C source to net sink and lost more C than DBF, suggesting ENF are likely slower to recover C assimilation capacity after stand-replacing harvests due to slower development of photosynthesis with stand age. Model results indicated 8790 that increasing harvesting intensity would delay recovery of NEP after clearing, but had little effect on C dynamics during late succession. Further improvements in numerical process-based forest population dynamic models for predicting the effects of climate change and forest harvests are considered.
Agricultural and Forest Meteorology, 2008
a g r i c u l t u r a l a n d f o r e s t m e t e o r o l o g y 1 4 8 ( 2 0 0 8 ) 2 8 8 -3 0 8 a ... more a g r i c u l t u r a l a n d f o r e s t m e t e o r o l o g y 1 4 8 ( 2 0 0 8 ) 2 8 8 -3 0 8 a r t i c l e i n f o Keywords: Carbon cycle Eddy covariance Managed and natural ecosystems Regional upscaling a b s t r a c t Carbon dioxide fluxes were examined over the growing seasons of 2002 and 2003 from 14
Agricultural and Forest Meteorology, 2008
a g r i c u l t u r a l a n d f o r e s t m e t e o r o l o g y 1 4 8 ( 2 0 0 8 ) 8 2 1 -8 3 8
Journal of Geophysical Research, 2010
1] Disturbances are important for renewal of North American forests. Here we summarize more than ... more 1] Disturbances are important for renewal of North American forests. Here we summarize more than 180 site years of eddy covariance measurements of carbon dioxide flux made at forest chronosequences in North America. The disturbances included standreplacing fire (Alaska, Arizona, Manitoba, and Saskatchewan) and harvest (British Columbia, Florida, New Brunswick, Oregon, Quebec, Saskatchewan, and Wisconsin) events, insect infestations (gypsy moth, forest tent caterpillar, and mountain pine beetle), Hurricane Wilma, and silvicultural thinning (Arizona, California, and New Brunswick). Net ecosystem production (NEP) showed a carbon loss from all ecosystems following a stand-replacing disturbance, becoming a carbon sink by 20 years for all ecosystems and by 10 years for most. Maximum carbon losses following disturbance (g C m −2 y −1 ) ranged from 1270 in Florida to 200 in boreal ecosystems. Similarly, for forests less than 100 years old, maximum uptake (g C m −2 y −1 ) was 1180 in Florida mangroves and 210 in boreal ecosystems. More temperate forests had intermediate fluxes. Boreal ecosystems were relatively time invariant after 20 years, whereas western ecosystems tended to increase in carbon gain over time. This was driven mostly by gross photosynthetic production (GPP) because total ecosystem respiration (ER) and heterotrophic respiration were relatively invariant with age. GPP/ER was as low as 0.2 immediately following stand-replacing disturbance reaching a constant value of 1.2 after 20 years. NEP following insect defoliations and silvicultural thinning showed lesser changes than stand-replacing events, with decreases in the year of disturbance followed by rapid recovery. NEP decreased in a mangrove ecosystem following Hurricane Wilma because of a decrease in GPP and an increase in ER.
Agricultural and Forest Meteorology, 2008
a g r i c u l t u r a l a n d f o r e s t m e t e o r o l o g y 1 4 8 ( 2 0 0 8 ) 2 1 6 -2 3 0 a ... more a g r i c u l t u r a l a n d f o r e s t m e t e o r o l o g y 1 4 8 ( 2 0 0 8 ) 2 1 6 -2 3 0 a r t i c l e i n f o a v a i l a b l e a t w w w . s c i e n c e d i r e c t . c o m j o u r n a l h o m e p a g e : w w w . e l s e v i e r . c o m / l o c a t e / a g r f o r m e t
The relative duration of active and dormant seasons has a strong influence on ecosystem net carbo... more The relative duration of active and dormant seasons has a strong influence on ecosystem net carbon balance and its carbon uptake potential. While recognized as an important source of temporal and spatial variability, the seasonality of ecosystem carbon balance has not been studied explicitly, and still lacks standard terminology. In the current chapter, we apply a curve fitting procedure to define seasonal transitions in ecosystem gross productivity (GEP) and respiration (ER), and we show that the temporal changes in these two fluxes are not synchronous, and that the transition dates and rates of change vary both across sites and between years. Carbon uptake period (CUP), a common phenological metric, defined from ecosystem net carbon exchange (NEE), is related to these periods of activity, but the differential sensitivities of GEP and ER to environmental factors complicate the interpretation of variation in CUP alone. On a landscape scale, differences in stand age represent a major source of heterogeneity reflected in different flux capacities as well as microclimate. In the current study, we evaluate age-related differences in the phenological transitions of GEP and ER using hardwood and conifer chronosequences. While a significant portion of variability in GEP seasonality was explained with stand age, the influence of interannual climatic variability exceeded these, and was the predominant factor affecting ER seasonality. The length of the
Agricultural and Forest Meteorology, 2009
Harvest disturbance has substantial impacts on forest carbon (C) fluxes and stocks. The quantific... more Harvest disturbance has substantial impacts on forest carbon (C) fluxes and stocks. The quantification of these effects is essential for the better understanding of forest C dynamics and informing forest management in the context of global change. We used a process-based forest ecosystem model, PnET-CN, to evaluate how, and by what mechanisms, clear-cuts alter ecosystem C fluxes, aboveground C stocks (AGC), and leaf area index (LAI) in northern temperate forests. We compared C fluxes and stocks predicted by the model and observed at two chronosequences of eddy covariance flux sites for deciduous broadleaf forests (DBF) and evergreen needleleaf forests (ENF) in the Upper Midwest region of northern Wisconsin and Michigan, USA. The average normalized root mean square error (NRMSE) and the Willmott index of agreement (d) for carbon fluxes, LAI, and AGC in the two chronosequences were 20 % and 0.90, respectively. Simulated gross primary productivity (GPP) increased with stand age, reaching a maximum (1200-1500 g C m −2 yr −1 ) at 11-30 years of age, and leveled off thereafter (900-1000 g C m −2 yr −1 ). Simulated ecosystem respiration (ER) for both plant functional types (PFTs) was initially as high as 700-1000 g C m −2 yr −1 in the first or second year after harvesting, decreased with age (400-800 g C m −2 yr −1 ) before canopy closure at 10-25 years of age, and increased to 800-900 g C m −2 yr −1 with stand development after canopy recovery. Simulated net ecosystem productivity (NEP) for both PFTs was initially negative, with net C losses of 400-700 g C m −2 yr −1 for 6-17 years after clear-cuts, reaching peak values of 400-600 g C m −2 yr −1 at 14-29 years of age, and eventually stabilizing in mature forests (> 60 years old), with a weak C sink (100-200 g C m −2 yr −1 ). The decline of NEP with age was caused by the relative flattening of GPP and gradual increase of ER. ENF recovered more slowly from a net C source to a net sink, and lost more C than DBF. This suggests that in general ENF may be slower to recover to full C assimilation capacity after stand-replacing harvests, arising from the slower development of photosynthesis with stand age. Our model results indicated that increased harvesting intensity would delay the recovery of NEP after clear-cuts, but this had little effect on C dynamics during late succession. Future modeling studies of disturbance effects will benefit from the incorporation of forest population dynamics (e.g., regeneration and mortality) and relationships between age-related model parameters and state variables (e.g., LAI) into the model.
Turbulent fluxes of carbon, water vapor, and temperature were continuously measured above an upla... more Turbulent fluxes of carbon, water vapor, and temperature were continuously measured above an upland forest in north central Wisconsin during 1999 and 2000 using the eddy covariance method. Maple (Acer saccharum), basswood (Tilia americana), and green ash (Fraxinus pennsylvanica) species found in this forest also comprise a substantial portion of the landscape in the northern Great Lakes region and area, and it has been hypothesized that forests of this age (60-80 years) are responsible for net uptake of atmospheric CO 2 over North America. Mean CO 2 , water vapor, and temperature profile measurements were used to improve flux estimates during periods of low turbulence, and were effective for friction velocities (u * ) >0.3 m s À1 . Unique observations at this site included nighttime and early morning venting anomalies that seemed to originate from a seemingly homogenous area within the forest. These elevated NEE measurements, some as high as 80 mol m À2 s À1 , appeared in valid turbulent flux observations for hours at a time, and provided circumstantial evidence for preferential venting and/or existence of pooled CO 2 in low-lying areas. We observed that the forest was a moderate sink for atmospheric carbon, and cumulative NEE of CO 2 was estimated to be À334 g C m À2 year À1 during 2000. Sensitivity to low-turbulence flux corrections was very small (21 g C m À2 year À1 ), and discrepancies between annual estimates of NEE and NEP were similar to other sites. A normalized measure of ecosystem respiration, the free energy of activation, was presented and its seasonal variations were analyzed. Gross ecosystem production (GEP) was high (1165 g C m À2 year À1 ) and ecosystem respiration (ER) was low (817 g C m À2 year À1 ) in comparison to spatially integrated, landscape-scale observations from WLEF (914 and 1005 g C m À2 year À1 , respectively), a 477 m tower located 22 km to the northeast [Glob. Change Biol. 9 (2003) 1278]. Forest transpiration was responsible for most of www.elsevier.com/locate/agrformet Agricultural and Forest Meteorology 126 (2004) 271-295
Geophysical Research Letters, 2004
1] Future climate change is expected to affect ecosystematmosphere CO 2 exchange, particularly th... more 1] Future climate change is expected to affect ecosystematmosphere CO 2 exchange, particularly through the influence of temperature. To date, however, few studies have shown that differences in the response of net ecosystem CO 2 exchange (NEE) to temperature among ecosystems can be explained by differences in the photosynthetic and respiratory processes that compose NEE. Using a new nonparametric statistical model, we analyzed data from four forest ecosystems. We observed that differences among forests in their ability to assimilate CO 2 as a function of temperature were attributable to consistent differences in the temperature dependence of photosynthesis and respiration. This observation provides empirical validation of efforts to develop models of NEE from the first-principle relationships between photosynthetic and respiratory processes and climate. Our results also showed that models of seasonal dynamics in NEE that lack specific consideration of the temperature dependence of respiration and photosynthesis are likely to carry significant uncertainties.
... Hasson, AS, Tyndall, GS, and Orlando, JJ: A product yield study of the reaction of HO2 radica... more ... Hasson, AS, Tyndall, GS, and Orlando, JJ: A product yield study of the reaction of HO2 radicals with ethyl peroxy (C2H5O2), acetyl peroxy (CH3C(O)O2), and acetonyl peroxy (CH3C(O)CH2O2) radicals, Journal of Physical Chemistry A, 108, 59795989, doi:10.1021/Jp048873t ...
Evapotranspiration (ET) prediction and forecasting play a vital role in improving water use in ag... more Evapotranspiration (ET) prediction and forecasting play a vital role in improving water use in agriculturally intensive areas. Metrological and biophysical predictors that drive ET in managed landscapes have complex nonlinear relationships. Deep learning and data-driven methods have shown promising performance for identifying the dependencies among variables. Here, we evaluated the potentials of random forest (RF) and long short-term memory (LSTM) neural networks to estimate and forecast daily ET for corn, soybeans, and potatoes in diverse agricultural farms during 2003-2019. The modeling framework was applied for nineteen fields where eddy covariance ET and meteorological observations in the Midwest USA for growing season (April-October) is available. In this study, we applied data-driven models (RF and LSTM) with 3 sets of predictors (5, 11, and 16 predictors). Results show that a 16 predictor RF model (RF_16 R 2 = 0.7, Willmott's skill score = 0.90) outperformed a process-bas...
The potential of a stepwise fusion of proximally sensed portable X-ray fluorescence (pXRF) spectr... more The potential of a stepwise fusion of proximally sensed portable X-ray fluorescence (pXRF) spectra and electromagnetic induction (EMI) with remote Sentinel-2 bands and a digital elevation model (DEM) was investigated for predicting soil physicochemical properties in pedons and across a heterogeneous 80-ha crop field in Wisconsin, USA. We found that pXRF spectra with partial least squares regression (PLSR) models can predict sand, total nitrogen (TN), organic carbon (OC), silt contents, and clay with validation R 2 of 0.81, 0.74, 0.73, 0.68, and 0.64 at the pedon scale but performed less well for soil pH (R 2 = 0.51). A combination of EMI, Sentinel-2, and DEM data showed promise in mapping sand, silt contents, and TN at two depths and Ap horizon thickness and soil depth across the field. A clustering analysis using combinations of mapped soil properties or proximal and remote sensing data suggested that data fusion improved the characterization of field-scale variability of soil properties. The cost-benefit analysis showed that the most accurate management zones (MZs) for topsoil can be generated only using estimated soil property maps while it was the most costly as compared to other data sources. For an intermediate-high (for topsoil) and high (subsoil) accuracy and a moderate economic budget, the combination of sensors (proximal + remote sensing + DEM) might be a better approach for effective MZs generation than collecting soil samples for laboratory analysis while the latter produced the most accurate maps for topsoil. It can be concluded that pXRF spectra can be useful for predicting key soil properties (e.g., sand, TN, OC, silt, clay) at different soil depths, and a combination of proximal and remote sensing provides an effective way to delineate soil MZs that are useful for decision-making.
Evapotranspiration (ET) prediction and forecasting play a vital role in improving water use in ag... more Evapotranspiration (ET) prediction and forecasting play a vital role in improving water use in agriculturally intensive areas. Metrological and biophysical predictors that drive ET in managed landscapes have complex nonlinear relationships. Deep learning and data-driven methods have shown promising performance for identifying the dependencies among variables. Here, we evaluated the potentials of random forest (RF) and long short-term memory (LSTM) neural networks to estimate and forecast daily ET for corn, soybeans, and potatoes in diverse agricultural farms during 2003-2019. The modeling framework was applied for nineteen fields where eddy covariance ET and meteorological observations in the Midwest USA for growing season (April-October) is available. In this study, we applied data-driven models (RF and LSTM) with 3 sets of predictors (5, 11, and 16 predictors). Results show that a 16 predictor RF model (RF_16 R 2 = 0.7, Willmott's skill score = 0.90) outperformed a process-based land surface model (LSM R 2 = 0.57, Willmott's skill score = 0.86) for predicting daily ET, while LSTM performance was lower (LSTM_16 R 2 = 0.65, Willmott's skill score = 0.89 and LSTM_11 R 2 = 0.62, Willmott's skill score = 0.86) than RF using the same sets of predictors. Vapor pressure and crop coefficients were identified as the most important predictors for irrigated crops, while short wave radiation and enhanced vegetation index were key predictors for non-irrigated crops. For certain crop types, such as corn and soybeans on fine-grained soils (silt loam), a simpler version RF, using only 11 drivers, can provide comparable results (R 2 = 0.70 vs 0.69 and Willmott's skill score = 0.90 vs 0.88). For short-term 3-day ET forecasting, LSTM is more sensitive to uncertainty in ensemble forecast meteorology than RF. ET forecasts were strongly sensitive to forecast uncertainty of vapor pressure. The proposed modeling architecture provides a field-scale, locally calibrated tool for accurate prediction and short-term forecasting of daily ET in areas where in situ ET, metrological, and biophysical data are lacking.
Sustainability, 2020
Dairy farms are predominantly carbon sources, due to high livestock emissions from enteric fermen... more Dairy farms are predominantly carbon sources, due to high livestock emissions from enteric fermentation and manure. Integrated crop-livestock systems (ICLSs) have the potential to offset these greenhouse gas (GHG) emissions, as recycling products within the farm boundaries is prioritized. Here, we quantify seasonal and annual greenhouse gas budgets of an ICLS dairy farm in Wisconsin USA using satellite remote sensing to estimate vegetation net primary productivity (NPP) and Intergovernmental Panel on Climate Change (IPCC) guidelines to calculate farm emissions. Remotely sensed annual vegetation NPP correlated well with farm harvest NPP (R 2 = 0.9). As a whole, the farm was a large carbon sink, owing to natural vegetation carbon sinks and harvest products staying within the farm boundaries. Dairy cows accounted for 80% of all emissions as their feed intake dominated farm feed supply. Manure emissions (15%) were low because manure spreading was frequent throughout the year. In combination with soil conservation practices, ICLS farming provides a sustainable means of producing nutritionally valuable food while contributing to sequestration of atmospheric CO2. Here, we introduce a simple and cost-efficient way to quantify whole-farm GHG budgets, which can be used by farmers to understand their carbon footprint, and therefore may encourage management strategies to improve agricultural sustainability.
Agricultural and Forest Meteorology, 2007
We review 15 techniques for estimating missing values of net ecosystem CO 2 exchange (NEE) in edd... more We review 15 techniques for estimating missing values of net ecosystem CO 2 exchange (NEE) in eddy covariance time series and evaluate their performance for different artificial gap scenarios based on a set of 10 benchmark datasets from six forested sites in Europe.
Stand-replacing disturbances including harvests have substantial impacts on forest carbon (C) flu... more Stand-replacing disturbances including harvests have substantial impacts on forest carbon (C) fluxes and stocks. The quantification and simulation of these effects is essential for better understanding forest C dynamics and informing forest management in the context of global change. We evaluated the process-based forest ecosystem 5 model, PnET-CN, for how well and by what mechanisms changes of ecosystem C fluxes, aboveground C stocks (AGC), and leaf area index (LAI) arise after clearcuts. We compared the effects of stand-replacing harvesting on C fluxes and stocks using two chronosequences of eddy covariance flux sites for deciduous broadleaf forests (DBF) and evergreen needleleaf forests (ENF) in the Upper Midwest region of northern 10 20 covery. Simulated NEP for both forest types was initially negative with the net C losses of ∼ 400-700 g C m −2 yr −1 for 6-17 years after harvesting, reached the peak values of ∼ 400-600 g C m −2 yr −1 at 14-29 years of age, and became stable and a weak C sink (∼ 100-200 g C m −2 yr −1 ) in mature forests (> 60 years old). The decline of NEP with age was caused by the relative flatting of GPP and gradual increasing of ER. ENF re-25 covered slower from net C source to net sink and lost more C than DBF, suggesting ENF are likely slower to recover C assimilation capacity after stand-replacing harvests due to slower development of photosynthesis with stand age. Model results indicated 8790 that increasing harvesting intensity would delay recovery of NEP after clearing, but had little effect on C dynamics during late succession. Further improvements in numerical process-based forest population dynamic models for predicting the effects of climate change and forest harvests are considered.
Agricultural and Forest Meteorology, 2008
a g r i c u l t u r a l a n d f o r e s t m e t e o r o l o g y 1 4 8 ( 2 0 0 8 ) 2 8 8 -3 0 8 a ... more a g r i c u l t u r a l a n d f o r e s t m e t e o r o l o g y 1 4 8 ( 2 0 0 8 ) 2 8 8 -3 0 8 a r t i c l e i n f o Keywords: Carbon cycle Eddy covariance Managed and natural ecosystems Regional upscaling a b s t r a c t Carbon dioxide fluxes were examined over the growing seasons of 2002 and 2003 from 14
Agricultural and Forest Meteorology, 2008
a g r i c u l t u r a l a n d f o r e s t m e t e o r o l o g y 1 4 8 ( 2 0 0 8 ) 8 2 1 -8 3 8
Journal of Geophysical Research, 2010
1] Disturbances are important for renewal of North American forests. Here we summarize more than ... more 1] Disturbances are important for renewal of North American forests. Here we summarize more than 180 site years of eddy covariance measurements of carbon dioxide flux made at forest chronosequences in North America. The disturbances included standreplacing fire (Alaska, Arizona, Manitoba, and Saskatchewan) and harvest (British Columbia, Florida, New Brunswick, Oregon, Quebec, Saskatchewan, and Wisconsin) events, insect infestations (gypsy moth, forest tent caterpillar, and mountain pine beetle), Hurricane Wilma, and silvicultural thinning (Arizona, California, and New Brunswick). Net ecosystem production (NEP) showed a carbon loss from all ecosystems following a stand-replacing disturbance, becoming a carbon sink by 20 years for all ecosystems and by 10 years for most. Maximum carbon losses following disturbance (g C m −2 y −1 ) ranged from 1270 in Florida to 200 in boreal ecosystems. Similarly, for forests less than 100 years old, maximum uptake (g C m −2 y −1 ) was 1180 in Florida mangroves and 210 in boreal ecosystems. More temperate forests had intermediate fluxes. Boreal ecosystems were relatively time invariant after 20 years, whereas western ecosystems tended to increase in carbon gain over time. This was driven mostly by gross photosynthetic production (GPP) because total ecosystem respiration (ER) and heterotrophic respiration were relatively invariant with age. GPP/ER was as low as 0.2 immediately following stand-replacing disturbance reaching a constant value of 1.2 after 20 years. NEP following insect defoliations and silvicultural thinning showed lesser changes than stand-replacing events, with decreases in the year of disturbance followed by rapid recovery. NEP decreased in a mangrove ecosystem following Hurricane Wilma because of a decrease in GPP and an increase in ER.
Agricultural and Forest Meteorology, 2008
a g r i c u l t u r a l a n d f o r e s t m e t e o r o l o g y 1 4 8 ( 2 0 0 8 ) 2 1 6 -2 3 0 a ... more a g r i c u l t u r a l a n d f o r e s t m e t e o r o l o g y 1 4 8 ( 2 0 0 8 ) 2 1 6 -2 3 0 a r t i c l e i n f o a v a i l a b l e a t w w w . s c i e n c e d i r e c t . c o m j o u r n a l h o m e p a g e : w w w . e l s e v i e r . c o m / l o c a t e / a g r f o r m e t
The relative duration of active and dormant seasons has a strong influence on ecosystem net carbo... more The relative duration of active and dormant seasons has a strong influence on ecosystem net carbon balance and its carbon uptake potential. While recognized as an important source of temporal and spatial variability, the seasonality of ecosystem carbon balance has not been studied explicitly, and still lacks standard terminology. In the current chapter, we apply a curve fitting procedure to define seasonal transitions in ecosystem gross productivity (GEP) and respiration (ER), and we show that the temporal changes in these two fluxes are not synchronous, and that the transition dates and rates of change vary both across sites and between years. Carbon uptake period (CUP), a common phenological metric, defined from ecosystem net carbon exchange (NEE), is related to these periods of activity, but the differential sensitivities of GEP and ER to environmental factors complicate the interpretation of variation in CUP alone. On a landscape scale, differences in stand age represent a major source of heterogeneity reflected in different flux capacities as well as microclimate. In the current study, we evaluate age-related differences in the phenological transitions of GEP and ER using hardwood and conifer chronosequences. While a significant portion of variability in GEP seasonality was explained with stand age, the influence of interannual climatic variability exceeded these, and was the predominant factor affecting ER seasonality. The length of the
Agricultural and Forest Meteorology, 2009
Harvest disturbance has substantial impacts on forest carbon (C) fluxes and stocks. The quantific... more Harvest disturbance has substantial impacts on forest carbon (C) fluxes and stocks. The quantification of these effects is essential for the better understanding of forest C dynamics and informing forest management in the context of global change. We used a process-based forest ecosystem model, PnET-CN, to evaluate how, and by what mechanisms, clear-cuts alter ecosystem C fluxes, aboveground C stocks (AGC), and leaf area index (LAI) in northern temperate forests. We compared C fluxes and stocks predicted by the model and observed at two chronosequences of eddy covariance flux sites for deciduous broadleaf forests (DBF) and evergreen needleleaf forests (ENF) in the Upper Midwest region of northern Wisconsin and Michigan, USA. The average normalized root mean square error (NRMSE) and the Willmott index of agreement (d) for carbon fluxes, LAI, and AGC in the two chronosequences were 20 % and 0.90, respectively. Simulated gross primary productivity (GPP) increased with stand age, reaching a maximum (1200-1500 g C m −2 yr −1 ) at 11-30 years of age, and leveled off thereafter (900-1000 g C m −2 yr −1 ). Simulated ecosystem respiration (ER) for both plant functional types (PFTs) was initially as high as 700-1000 g C m −2 yr −1 in the first or second year after harvesting, decreased with age (400-800 g C m −2 yr −1 ) before canopy closure at 10-25 years of age, and increased to 800-900 g C m −2 yr −1 with stand development after canopy recovery. Simulated net ecosystem productivity (NEP) for both PFTs was initially negative, with net C losses of 400-700 g C m −2 yr −1 for 6-17 years after clear-cuts, reaching peak values of 400-600 g C m −2 yr −1 at 14-29 years of age, and eventually stabilizing in mature forests (> 60 years old), with a weak C sink (100-200 g C m −2 yr −1 ). The decline of NEP with age was caused by the relative flattening of GPP and gradual increase of ER. ENF recovered more slowly from a net C source to a net sink, and lost more C than DBF. This suggests that in general ENF may be slower to recover to full C assimilation capacity after stand-replacing harvests, arising from the slower development of photosynthesis with stand age. Our model results indicated that increased harvesting intensity would delay the recovery of NEP after clear-cuts, but this had little effect on C dynamics during late succession. Future modeling studies of disturbance effects will benefit from the incorporation of forest population dynamics (e.g., regeneration and mortality) and relationships between age-related model parameters and state variables (e.g., LAI) into the model.
Turbulent fluxes of carbon, water vapor, and temperature were continuously measured above an upla... more Turbulent fluxes of carbon, water vapor, and temperature were continuously measured above an upland forest in north central Wisconsin during 1999 and 2000 using the eddy covariance method. Maple (Acer saccharum), basswood (Tilia americana), and green ash (Fraxinus pennsylvanica) species found in this forest also comprise a substantial portion of the landscape in the northern Great Lakes region and area, and it has been hypothesized that forests of this age (60-80 years) are responsible for net uptake of atmospheric CO 2 over North America. Mean CO 2 , water vapor, and temperature profile measurements were used to improve flux estimates during periods of low turbulence, and were effective for friction velocities (u * ) >0.3 m s À1 . Unique observations at this site included nighttime and early morning venting anomalies that seemed to originate from a seemingly homogenous area within the forest. These elevated NEE measurements, some as high as 80 mol m À2 s À1 , appeared in valid turbulent flux observations for hours at a time, and provided circumstantial evidence for preferential venting and/or existence of pooled CO 2 in low-lying areas. We observed that the forest was a moderate sink for atmospheric carbon, and cumulative NEE of CO 2 was estimated to be À334 g C m À2 year À1 during 2000. Sensitivity to low-turbulence flux corrections was very small (21 g C m À2 year À1 ), and discrepancies between annual estimates of NEE and NEP were similar to other sites. A normalized measure of ecosystem respiration, the free energy of activation, was presented and its seasonal variations were analyzed. Gross ecosystem production (GEP) was high (1165 g C m À2 year À1 ) and ecosystem respiration (ER) was low (817 g C m À2 year À1 ) in comparison to spatially integrated, landscape-scale observations from WLEF (914 and 1005 g C m À2 year À1 , respectively), a 477 m tower located 22 km to the northeast [Glob. Change Biol. 9 (2003) 1278]. Forest transpiration was responsible for most of www.elsevier.com/locate/agrformet Agricultural and Forest Meteorology 126 (2004) 271-295
Geophysical Research Letters, 2004
1] Future climate change is expected to affect ecosystematmosphere CO 2 exchange, particularly th... more 1] Future climate change is expected to affect ecosystematmosphere CO 2 exchange, particularly through the influence of temperature. To date, however, few studies have shown that differences in the response of net ecosystem CO 2 exchange (NEE) to temperature among ecosystems can be explained by differences in the photosynthetic and respiratory processes that compose NEE. Using a new nonparametric statistical model, we analyzed data from four forest ecosystems. We observed that differences among forests in their ability to assimilate CO 2 as a function of temperature were attributable to consistent differences in the temperature dependence of photosynthesis and respiration. This observation provides empirical validation of efforts to develop models of NEE from the first-principle relationships between photosynthetic and respiratory processes and climate. Our results also showed that models of seasonal dynamics in NEE that lack specific consideration of the temperature dependence of respiration and photosynthesis are likely to carry significant uncertainties.
... Hasson, AS, Tyndall, GS, and Orlando, JJ: A product yield study of the reaction of HO2 radica... more ... Hasson, AS, Tyndall, GS, and Orlando, JJ: A product yield study of the reaction of HO2 radicals with ethyl peroxy (C2H5O2), acetyl peroxy (CH3C(O)O2), and acetonyl peroxy (CH3C(O)CH2O2) radicals, Journal of Physical Chemistry A, 108, 59795989, doi:10.1021/Jp048873t ...
The Central Sands region of Wisconsin is characterized by productive trout streams, lakes, farmla... more The Central Sands region of Wisconsin is characterized by productive trout streams, lakes, farmland and forest. However, stream channelization, past wetland drainage, and ground water withdrawals have disrupted the hydrology of this Central Sands region. Climatically driven conditions in last decade (2000-2008) alone are unable to account for the severely depressed water levels. Increased interception and evapotranspiration from afforested areas in central sand Wisconsin may also be culprit for reduced water recharge. Hence, there is need to study the cumulative effects of changing precipitation patterns, groundwater withdrawals, and forest evapotranspiration to improve projections of the future of lake levels and water availability in this region. Here, the SWAT-MODFLOW coupled model approach was applied at large spatio-temporal scale. The coupled model fully integrates a watershed model (SWAT) with a groundwater flow model (MODFLOW). Surface water and ground water flows were simulated integratively at daily time step to estimate the groundwater discharge to the stream network in Central Sands that encompasses high capacity wells. The model was calibrated (2010-2013) and validated (2014-2017) based on streamflow, groundwater extraction, and water table elevation. As the long-term trends in some of the primary drivers is presently ambiguous in Central Sands under future climate, as is the case for total precipitation or timing of precipitation, we relied on a sensitivity student to quantitatively access how primary and secondary drivers may influence future net groundwater recharge. We demonstrate how such an approach could then be coupled with decision-making models to evaluate the effectiveness of groundwater withdrawal policies under a changing climate.