"First - I am strongly supportive of this effort. The committee that drafted this white paper represents some of the best savanna ecologists in the world - so congratulations on a job well done. The primary input I have would be to seriously consider adding a distributed, coordinated field experiment in these distributed sites around the world. When assessing global phenomenon such as shrub encroachment, a major limitation is that field studies and critical experiments are all conducted independently and with different methods, at varying spatial scales and they measure different response variables. This weakens our abilities to fully utilize remote sensing data and models. There are clever ways to conduct distributed meta-experiments (e.g., the NutNet example) that can be transformative. I would be happy participate in and support such an addition to this proposed program. "

"I think this proposal will improve our knowledge in savanna (tree-grass) ecosystem. I have some comments on the proposal. It seems that observation and modeling of grass and tree seasonality (phenology) were not well mentioned on the proposal. If annual grass and deciduous trees coexist, the estimation of biophysical parameters, such as LAI, from satellite will be challenging. I would expect that this project contribute to answer how we get the grass and tree parameters separately via remote sensing techniques, and how we can utilize such information in the models proposed. Current satellite products based on MODIS are not sufficient for the analysis of tree-grass ecosystems. To implement this, some theoretical investigations based on radiative transfer simulation will be necessary. The airborne and satellite LiDAR data are very important. NASA's program has been postponed, but I heard that some Japanese scientists are thinking of installing the canopy LiDAR on the International Space Station. It could be useful for the tree-grass ecosystem researches. "

"The case for a savanna focussed effort has become stronger in the context of REDD+ in terms of the UNFCCC program. Much of the dynamism of the terrestrial part of teh global C cycle is traceable to mixed tree-grass systems, rather than to the more stable forsts systems (in the narrow sense. The remote sensing challenges are quite different - structure becomes important, and detection of gradual change rather than land cover transitions."

"To the extent that global changes in rainfall patters and amounts will effect vegetation, monitoring tree-grass savannas is worthwhile goal. However, the white paper outlines a great deal of process level work that is extremely expensive and will likely to be very limited in extent. This situation makes generalization of the results with models questionable at the global scale. Several recent presentations & papers are calling into question the predictability of these models as well as a recent presentation by Steve Pacala at ESA this summer on prediction reliability in well calibrated ecological models. My suggestion is that you make a realistic break in scaling such that the remote sensing data can be linked to more robust land use, and structure and function components that can be validated globally. Along those lines, my suggestion is also to tie into savanna based research sites with established research, particularly well characterized landscapes to leaverage their studies. For example, the Savannah River Site in SC has implemented millions of dollars in NSF, SERDP, JFSP funds on savanna related research. There are also several biological field stations not listed in the Appendix that would be good partners."

"1. The earlier passages leave science parts last in lists of aims, resulting in somewhat unbalanced prominence of remote sensing technique development. Obviously technique development is essential to address the science issues, but the science issues come first. In Section 2, the sequence is correct. 2. As the White Paper states, savannas have enormous global significance but with, to some extent, strong regional differences. The NASA contribution to the global component is rather tentative in the White Paper (p.20). I recommend that NASA funds USA science proposals to work in the major savanna types world-wide; and ii. high-level negotiations with international (e.g. AMMA, Carboafrica, FAO) and country (to add to existing strong programs in, for example, Australia and South Africa)research programs are started immediately a decision to proceed is made. "

"It should be emphasized somewhere in the text that the savanna ecosystem is not only sensitive to climate change and land-use change as a whole system, but also the unique interaction between distinct grass and trees within the community make it more sensitive than other ecosystems. For example, the phenology shift in response to warming could be different between grass and trees, resulting in a potential mismatch of ecosystem structure and function. This could be reflected in high-resolution remote sensing data and simulated carbon models. "

"Overall the white paper identifies an important science need and proposes an impressive suite of measurements to better understand tree-grass ecosystems. One issue that could be highlighted more strongly is active conversion of forests to mixed tree-grass systems via tree thinning and prescribed burning treatments to reduce fuels and reduce wildfire intensity and hazard. This conversion is currently common and will be increasing in the future in ponderosa-pine-dominated forests of the southwestern US and southern Rocky Mountains. For example, these treatments are planned for about 2 million acres of forests in northern Arizona. Thinning the trees greatly increases grass production. Thus, the science question is: how does creation of savanna-like tree-grass systems from current forests via restoration activities affect carbon, water, and energy balances, and climate feedbacks. I manage two ameriflux eddy covariance sites in northern Arizona that might be suitable for the T-G initiative: flagstaff-managed and flagstaff-wildfire. The flagstaff-managed site is a ponderosa pine forest that was recently thinned to create a mixed tree-grass system. The flagstaff-wildfire site was converted from forest to grass/shrubland by an intense fire 15 years ago. Both are described in Dore et al. (2010). Carbon and water fluxes from ponderosa pine forests disturbed by wildfire and thinning. Ecological Applications 20(3):663-683."

"The Tree-Grass white paper is a critically important document that lays out the multiple scientific and societal motivations for a NASA field, modeling, and remote sensing campaign focused on earth's savanna ecosystems. This is a generally excellent and visionary document, and I enthusiastically endorse the overall goals it articulates. My comments below focus on one area I think should be more fully addressed in planning for a campaign: the profound differences in physiology and function between C4 grasses and other savanna plants (both herbaceous and woody) that use the C3 photosynthetic pathway. Although the white paper mentions that C4 grasses largely dominate the continuous herbaceous understory of most savannas, I would like to amplify the importance of these grasses in savanna structure and function and in the response of savannas to fires, climate variation, and rising atmospheric CO2. Indeed, as noted in the white paper, the emergence of the savanna biome was largely driven by the dramatic expansion of C4 grasses in the late Miocene, a phenomenon that is largely unexplained, but likely explanations include climate drying, fire, and disturbance. There is considerable theoretical and experimental evidence that many savannas have evolved adaptations to frequent fires. Indeed, fire has been proposed as a primary factor in determining tropical and subtropical savanna structure, composition and distribution; such savannas are often dominated (in terms of cover and productivity, if not species diversity) by understory grasses that use the C4 photosynthetic pathway. Indeed, many C4 grass species are thought to be dependent on regular fires for their maintenance and growth: C4 grasses typically thrive in open environments that experience frequent disturbance and high precipitation variability. C4 plants have higher photosynthetic rates at high temperatures and under high light conditions. C4 plants exhibit higher light-use and water-use efficiency than comparable C3 plants, and they also are expected to respond very differently to rising CO2. Finally, C4 plants fractionate atmospheric 13CO2 to a lesser extent than do C3 plants, and thus carbon fluxes from savannas can be partitioned into C3 (forbs, shrubs, trees) and C4 grass components using atmospheric 13CO2 measurements. Such measurements are being revolutionized by laser spectroscopy systems. In addition to effects of the C4 photosynthetic pathway, differences imposed by growth form and evolutionary history between understory herbaceous grasses and woody shrubs and trees in savanna ecosystems also produce fundamental ecological differences that will determine how these plant functional types respond to climate change. Savannas are also critically important in the global carbon cycle, as inferred from recent inversion models estimates. Similarly, land surface models simulate particularly high interannual variability of photosynthesis by savanna ecosystems, and this is partly driven by the differing functional responses of C4 grasses. Remote sensing of savannas and tree:grass systems represents a frontier challenge. In my own work estimating the C4 grass fraction in savannas, a critical limitation is the inability of products like the MODIS VCF to distinguish low-stature woody vegetation (e.g., shrubs) from herbaceous grassy vegetation. This is a critical challenge that future missions should address. Satellite observations of fire events facilitate studies of the spatial distribution and frequency of fire activity, and yet few studies have explored relationships among the C4 vegetation fraction, climate, and fire frequency. Such observations will contribute to the evolving understanding of how fires may maintain C4 grass-dominated savannas, especially those with sufficient precipitation to support tree growth. Satellite observations of other aspects of savanna form and function, such as phenology (e.g., EVI) and canopy temperature (e.g., LST), will also critically enhance our understanding of the savanna biome. Field and modeling campaigns such as are envisioned in the Tree-Grass white paper would usefully complement such remote sensing."

"Given the large area and little attention of the Tree-Grass ecosystem in the past, the proposed Field Campaign is very necessary. The existing national database such as National Land Cover Dataset, has classified shrub/scrub as areas with shrub canopy greater than 20%. However, this lower estimate the actual carbon stored in shrub area because many shrub or tree-grass ecosystems may have less than 20% canopy cover, such as sagebrush steppe in Western USA. Hopefully the Field Campaign can produce more accurate mapping method and create better land cover classification for this type of ecosystem. The existing ecosystem level biogeochemical models are mainly developed for single ecosystem type. Though some models, like CENTURY model, have considered interactions between Tree and Grass (e.g. tree canopy cover effect on grassland production, the nutrient competition between tree and grass), further test and development of the model in tree grass system will be needed. This Campaign can provide a good opportunity to test the scientific hypothesis in these models and promote our understanding in biogeochemical cycles. The research will also benefit other projects that related to the Tree-Grass ecosystem such as the Wildland Fire study in 2011 NASA ROSES, USGS NLCD and USGS's National Assessment of Ecological Carbon Stocks, Carbon Sequestration, and Greenhouse Gas Fluxes. Other federal agencies such as BLM, FWS and NRCS, may also contribute to and benefit from this Field Campaign. For example, the recent BLM National Greater Sage-Grouse Land Use Planning Strategy may get better land use planning of Greater Sage-Grouse habitat with more accurate shrubland mapping and change detection. I have commented on the draft of the plan and would like to see this Field Campaign to move forward. I also would likely work on a proposal for this call on model development."

"Thank you for the opportunity to comment on the Tree Grass proposal. I must agree with the argument that this work is valuable and necessary. After all, intelligent management of our resources is vital to our survival. An inherent advantage to the Tree Grass proposal is the location of its primary sites. By conducting the bulk of the intensive field campaigns within the USA, rather than abroad, investigators will enjoy lower operational costs and access to vast quantities of validated information. It is my recommendation, that the investigators and their sponsors promote environmental stewardship, while planning and conducting their investigations. For example, in lieu of placing a conventional trailer at a desirable field site, perhaps a green building can be erected. A sustainable practice, such as this one, can serve educational purposes and generate goodwill, when that building is later used as a research lab and/or nature center. It would be my pleasure to assist with the Tree Grass initiative."

"The 'tree-grass' map shown in Fig. 1 obviously includes the agricultural areas. Will this proposed project study the croplands? I think the croplands are different from tree-grass system in both their ecosystem functions and relations to climate change. NASA's FIFE (the First International Satellite Land Surface Climatology Project) covers an area with a good portion of grassland, and BOREAS (the Boreal Ecosystem-Atmosphere Study) and LBA (Large-scale Biosphere-Atmosphere Experiment) covered forested areas. Probably it is a time to have a NASA project to investigate the land-atmosphere interaction at tree-grass region, especially because it is an ecosystem providing significant goods and service to human and is vulnerable to climate change and human disturbances. We have accomplished a lot in remote sensing technology and modeling (especially the data assimilation) since FIFE and the tree-grass project should be based on the lessons learned from previous project and have very clear objectives and sound experiment plan. One of the recommendations from the workshop (March 2-4, 2010) was An appreciation of current capacity for measurement, remote sensing retrieval, modeling and prediction in tree-grass systems must be obtained to more clearly identify deficiencies, needs, and developments to be targeted by an intensive field campaign. The challenges & opportunities listed in Table 2 do not mention the scale issue. High resolution spectral image and lidar scanner data will be able to measure the structures of tree-grass system. But for resolutions of MODIS data, what else we can do in addition to the MODIS VCF data and other MODIS products? The parameters listed in Table 4 are comprehensive, but not specific to tree-grass region. What are the major issues need to be addressed by NASA's remote sensing technology? What are the problems in modeling? They should be identified to justify a large scale experiment. "