Hi Adrian, this is a great question, and one I'm pretty familiar with. Taking it from the 'top of the catchment' through to the 'final sink' assuming that it's a seaward draining catchment I'd run with the following overview:

1. For hillslope erosion, you can use the standard ULSE or RUSLE type equations and use a remotely sensed fractional cover estimate such as that suggested by Guerschman et al

http://www.sciencedirect.com/science/article/pii/S0034425709000121

This provides the capacity to estimate the ground cover over a period of time or for a specific runoff event. From there you can estimate the hillslope erosion term in your sediment budget. For smaller catchments the approach described in Guerschman et al can be applied to Landsat data.

2. For gully erosion you'll need to either model/estimate it or you can fork out for multi-temporal high 50cm increase in depth... InSAR might provide answers too, but that's a significant investment in both imagery and image processing techniques.

5. If you're chasing end of catchment sediment budgets you can always look at multi-temporal Landsat or MODIS to check out plume dynamics, particularly during big flood events. You'll need a locally tuned enrichment ratio to estimate the coarse (rather than suspended) fractions and have a friendly estuarine modeller to model your mixing floculation dynamics.

6. And all of the above depend on a decent DEM to drive all of your process models. SRTM is your best option here provided you can strip the trees out. Most 'stereo' based DEMS don't provide sufficient resolution on low relief areas such as floodplains etc.

YOu'll likely need to mix and match these approaches depending on the erosion processes that dominate in your catchment and the budget you have to hand...

Have fun!

Hi Adrian, this is a great question, and one I'm pretty familiar with. Taking it from the 'top of the catchment' through to the 'final sink' assuming that it's a seaward draining catchment I'd run with the following overview:

1. For hillslope erosion, you can use the standard ULSE or RUSLE type equations and use a remotely sensed fractional cover estimate such as that suggested by Guerschman et al

http://www.sciencedirect.com/science/article/pii/S0034425709000121

This provides the capacity to estimate the ground cover over a period of time or for a specific runoff event. From there you can estimate the hillslope erosion term in your sediment budget. For smaller catchments the approach described in Guerschman et al can be applied to Landsat data.

2. For gully erosion you'll need to either model/estimate it or you can fork out for multi-temporal high 50cm increase in depth... InSAR might provide answers too, but that's a significant investment in both imagery and image processing techniques.

5. If you're chasing end of catchment sediment budgets you can always look at multi-temporal Landsat or MODIS to check out plume dynamics, particularly during big flood events. You'll need a locally tuned enrichment ratio to estimate the coarse (rather than suspended) fractions and have a friendly estuarine modeller to model your mixing floculation dynamics.

6. And all of the above depend on a decent DEM to drive all of your process models. SRTM is your best option here provided you can strip the trees out. Most 'stereo' based DEMS don't provide sufficient resolution on low relief areas such as floodplains etc.

YOu'll likely need to mix and match these approaches depending on the erosion processes that dominate in your catchment and the budget you have to hand...

Have fun!

Thanks for this elaborate answer, Leo. I think this provides a good starting point for getting deeper into the subject and maybe also for starting a fruitful discussion here! I'll try to get a deeper look into the articeles you refered to.

I know your question specifies satellite, but there have been some cool studies done using airborne LiDAR. This follows on a bit from Leo's fourth point, I suppose. Thoma et al (2005) studied mass wasting along a riverbank, and found that it was an effective erosion-monitoring tool. Sediment-tracking was also done by Baldo et al (2009) for studying mudflows in Italy. And while I haven't seen anything about it, there might be some application of bathymetric LiDAR in estimating water turbidity as a proxy of suspended sediment. I also know that proposals for satellite-based LiDAR systems are increasing in prevalence (e.g., Yu et al (2010), Simard et al (2011)), so there might be some more recent work in combining these ideas...

HI all - just came upon this discussion thread now- bit late - sorry - but thought I would add my 2 cents worth. As Leo and others have highlighted I think there are a range of remote sensing techniques that can contribute to the construction of a catchment sediment budget - but there is no magic bullet - and never will be. You have to rely on multiple data sources at a range of scales - built on a solid program of field data collection.

If you are interested - have at look at some of the things we have done in Cape York - http://www.capeyorkwaterquality.info/ and there are a bunch of papers and reports all available through ResearchGate

That being said I cant stress enough the value of repeat LiDAR - despite its expense. And also dont overlook good old fashioned aerial photography for understanding multi -decadal scale geomorphic changes.

I would urge some caution about the use of RUSLE based models for predicting hillslope erosion in certain areas. We have just submitted a couple of papers that show RUSLE was overpredicting hillslope erosion rates by up to 4 orders of magnitude in northern Australia. The problem is not so much with cover factor (and the fractional cover analysis certainly improves things a bit - but not by orders of magnitude) - but in the understanding of soil erodibility (K factor). Again - this highlights the need for backing all of these things up with a solid field program to validate any predictions based on models and remote sensed data. RUSLE is an empirical model and can only be corrected with field measurements from the area in question.

I should also add that all of these data should be complemented with geochemical tracing data.

Thanks again for all of your answers and contributions to the discussion. I hope that I will have some time to take a deeper look into some of the papers you suggested during Christmas holidays.

So far it seems that mostly airborne and especially LiDAR aproaches are fruitful methods for assessing smaller scale aspects of catchment sediment budgets.

What I'm thinking about at the moment is how useful the upcoming Sentinel-1 and 2 Satellites could be in this respect. I know that there's already being done much work regarding the develpment of methods to use their data for example for lake monitoring or wetland monitoring and I'm curious if anyone of you knows of any activities or plans to use them for landcover/landuse, erosion or sediment budget monitoring?